modelpedia 0.0.3 → 0.0.5

package/dist/providers/openai.cjs

@@ -1,3 +1,3 @@
-Object.defineProperty(exports,Symbol.toStringTag,{value:`Module`});const e={id:`openai`,name:`OpenAI`,region:`US`,url:`https://openai.com`,api_url:`https://api.openai.com/v1`,docs_url:`https://platform.openai.com/docs`,pricing_url:`https://openai.com/api/pricing`,playground_url:`https://platform.openai.com/playground`,icon:`<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24" fill="currentColor" fill-rule="evenodd">
+Object.defineProperty(exports,Symbol.toStringTag,{value:`Module`});const e={id:`openai`,name:`OpenAI`,region:`US`,headquarters:`San Francisco, CA`,founded:2015,url:`https://openai.com`,api_url:`https://api.openai.com/v1`,docs_url:`https://platform.openai.com/docs`,pricing_url:`https://openai.com/api/pricing`,description:`Creator of GPT, DALL-E, and Sora. Builds frontier AI models for text, image, video, and audio.`,type:`direct`,playground_url:`https://platform.openai.com/playground`,status_url:`https://status.openai.com`,changelog_url:`https://platform.openai.com/docs/changelog`,sdk:{python:`openai`,javascript:`openai`},openai_compatible:!1,free_tier:!1,github_url:`https://github.com/openai`,models_url:`https://platform.openai.com/docs/models`,twitter_url:`https://x.com/openai`,discord_url:`https://discord.com/invite/openai`,blog_url:`https://openai.com/index`,terms_url:`https://openai.com/policies/terms-of-use`,support_url:`https://help.openai.com`,icon:`<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24" fill="currentColor" fill-rule="evenodd">
   <path d="M9.205 8.658v-2.26c0-.19.072-.333.238-.428l4.543-2.616c.619-.357 1.356-.523 2.117-.523 2.854 0 4.662 2.212 4.662 4.566 0 .167 0 .357-.024.547l-4.71-2.759a.797.797 0 00-.856 0l-5.97 3.473zm10.609 8.8V12.06c0-.333-.143-.57-.429-.737l-5.97-3.473 1.95-1.118a.433.433 0 01.476 0l4.543 2.617c1.309.76 2.189 2.378 2.189 3.948 0 1.808-1.07 3.473-2.76 4.163zM7.802 12.703l-1.95-1.142c-.167-.095-.239-.238-.239-.428V5.899c0-2.545 1.95-4.472 4.591-4.472 1 0 1.927.333 2.712.928L8.23 5.067c-.285.166-.428.404-.428.737v6.898zM12 15.128l-2.795-1.57v-3.33L12 8.658l2.795 1.57v3.33L12 15.128zm1.796 7.23c-1 0-1.927-.332-2.712-.927l4.686-2.712c.285-.166.428-.404.428-.737v-6.898l1.974 1.142c.167.095.238.238.238.428v5.233c0 2.545-1.974 4.472-4.614 4.472zm-5.637-5.303l-4.544-2.617c-1.308-.761-2.188-2.378-2.188-3.948A4.482 4.482 0 014.21 6.327v5.423c0 .333.143.571.428.738l5.947 3.449-1.95 1.118a.432.432 0 01-.476 0zm-.262 3.9c-2.688 0-4.662-2.021-4.662-4.519 0-.19.024-.38.047-.57l4.686 2.71c.286.167.571.167.856 0l5.97-3.448v2.26c0 .19-.07.333-.237.428l-4.543 2.616c-.619.357-1.356.523-2.117.523zm5.899 2.83a5.947 5.947 0 005.827-4.756C22.287 18.339 24 15.84 24 13.296c0-1.665-.713-3.282-1.998-4.448.119-.5.19-.999.19-1.498 0-3.401-2.759-5.947-5.946-5.947-.642 0-1.26.095-1.88.31A5.962 5.962 0 0010.205 0a5.947 5.947 0 00-5.827 4.757C1.713 5.447 0 7.945 0 10.49c0 1.666.713 3.283 1.998 4.448-.119.5-.19 1-.19 1.499 0 3.401 2.759 5.946 5.946 5.946.642 0 1.26-.095 1.88-.309a5.96 5.96 0 004.162 1.713z"/>
-</svg>`,models:[{id:`babbage-002`,name:`babbage-002`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`GPT base models can understand and generate natural language or code but are not trained with instruction following. These models are made to be replacements for our original GPT-3 base models and use the legacy Completions API. Most customers should use GPT-3.5 or GPT-4.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:16384,model_type:`chat`,performance:1,speed:3,tagline:`Replacement for the GPT-3 ada and babbage base models`,successor:`gpt-4o`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.4,1.6,null,1.6]},{label:`Batch`,values:[null,.8,null,.8]}]},{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[.4,.4]},{label:`Batch`,values:[.2,.2]}]}],input:.4,output:.4,batch_input:.2,batch_output:.2},endpoints:[`completions`],snapshots:[`babbage-002`]},{id:`chatgpt-4o-latest`,name:`ChatGPT-4o`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`chatgpt`,description:`ChatGPT-4o points to the GPT-4o snapshot currently used in ChatGPT. We recommend using an API model like [GPT-5](/api/docs/models/gpt-5) or [GPT-4o](/api/docs/models/gpt-4o) for most API integrations, but feel free to use this ChatGPT-4o model to test our latest improvements for chat use cases.`,status:`deprecated`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:3,speed:3,tagline:`GPT-4o model used in ChatGPT`,capabilities:{streaming:!0,vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[5,15]}]}],input:5,output:15},endpoints:[`chat_completions`,`responses`],snapshots:[`chatgpt-4o-latest`]},{id:`chatgpt-image-latest`,name:`chatgpt-image-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`chatgpt`,description:`GPT Image Latest points to the image snapshot currently used in ChatGPT.`,status:`active`,model_type:`image`,performance:5,speed:3,tagline:`Image model used in ChatGPT.`,modalities:{input:[`text`,`image`],output:[`image`,`text`]},pricing:{input:5,output:10,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,10]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,2,32]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.009,.013,.013]},{label:`Medium`,values:[null,.034,.05,.05]},{label:`High`,values:[null,.133,.2,.2]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`],snapshots:[`chatgpt-image-latest`]},{id:`Code Interpreter`,name:`Code Interpreter`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,pricing:{tiers:[{label:`Built-in tools`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.03]}]}]}},{id:`codex-mini-latest`,name:`codex-mini-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`codex`,description:`codex-mini-latest is a fine-tuned version of o4-mini specifically`,status:`deprecated`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Fast reasoning model optimized for the Codex CLI`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.5,output:6,cached_input:.375,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.5,.375,6]}]}]},endpoints:[`responses`],snapshots:[`codex-mini-latest`]},{id:`computer-use-preview-2025-03-11`,name:`computer-use-preview-2025-03-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`The computer-use-preview model is a specialized model for the computer use`,status:`active`,knowledge_cutoff:`2023-09`,context_window:8192,max_output_tokens:1024,reasoning_tokens:!0,alias:`computer-use-preview`,performance:2,reasoning:2,speed:2,tagline:`The computer-use-preview model is a specialized model for the computer use`,capabilities:{tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:3,output:12,batch_input:1.5,batch_output:6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,null,12]},{label:`Batch`,values:[1.5,null,6]}]}]},endpoints:[`responses`,`batch`]},{id:`computer-use-preview`,name:`computer-use-preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`The computer-use-preview model is a specialized model for the computer use`,status:`active`,knowledge_cutoff:`2023-09`,context_window:8192,max_output_tokens:1024,reasoning_tokens:!0,performance:2,reasoning:2,speed:2,tagline:`Specialized model for computer use tool`,capabilities:{tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:3,output:12,batch_input:1.5,batch_output:6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,null,12]},{label:`Batch`,values:[1.5,null,6]}]}]},endpoints:[`responses`,`batch`],snapshots:[`computer-use-preview-2025-03-11`]},{id:`dall-e-2`,name:`DALL·E 2`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`DALL·E is an AI system that creates realistic images and art from a natural language description. Older than DALL·E 3, DALL·E 2 offers more control in prompting and more requests at once.`,status:`deprecated`,model_type:`image`,performance:1,speed:2,tagline:`Our first image generation model`,successor:`dall-e-3`,modalities:{input:[`text`],output:[`image`]},pricing:{tiers:[{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`256x256`,`512x512`,`1024x1024`],rows:[{label:`Standard`,values:[null,.016,.018,.02]}]}]},endpoints:[`image_generation`,`image_edit`],snapshots:[`dall-e-2`]},{id:`dall-e-3`,name:`DALL·E 3`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`DALL·E is an AI system that creates realistic images and art from a natural language description. DALL·E 3 currently supports the ability, given a prompt, to create a new image with a specific size.`,status:`deprecated`,model_type:`image`,performance:3,speed:2,tagline:`Previous generation image generation model`,modalities:{input:[`text`],output:[`image`]},pricing:{tiers:[{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1792`,`1792x1024`],rows:[{label:`Standard`,values:[null,.04,.08,.08]},{label:`HD`,values:[null,.08,.12,.12]}]}]},endpoints:[`image_generation`],snapshots:[`dall-e-3`]},{id:`davinci-002`,name:`davinci-002`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`GPT base models can understand and generate natural language or code but are not trained with instruction following. These models are made to be replacements for our original GPT-3 base models and use the legacy Completions API. Most customers should use GPT-3.5 or GPT-4.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:16384,model_type:`chat`,performance:1,speed:3,tagline:`Replacement for the GPT-3 curie and davinci base models`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[6,12,null,12]},{label:`Batch`,values:[null,6,null,6]}]},{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[2,2]},{label:`Batch`,values:[1,1]}]}],input:2,output:2,batch_input:1,batch_output:1},endpoints:[`completions`],snapshots:[`davinci-002`]},{id:`File Search Storage`,name:`File Search Storage`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,pricing:{tiers:[{label:`Built-in tools`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.1]}]}]}},{id:`gpt-3.5-0301`,name:`gpt-3.5-0301`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-3.5`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,performance:1,speed:2,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`]},{id:`gpt-3.5-turbo-0125`,name:`gpt-3.5-turbo-0125`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-3.5`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,performance:1,speed:2,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`]},{id:`gpt-3.5-turbo-0613`,name:`gpt-3.5-turbo-0613`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-3.5`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,performance:1,speed:2,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`batch`]},{id:`gpt-3.5-turbo-16k-0613`,name:`gpt-3.5-turbo-16k-0613`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-3.5`,description:`GPT-3.5 Turbo models can understand and generate natural language or code and have been optimized for chat using the Chat Completions API but work well for non-chat tasks as well. As of July 2024, use gpt-4o-mini in place of GPT-3.5 Turbo, as it is cheaper, more capable, multimodal, and just as fast. GPT-3.5 Turbo is still available for use in the API.`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,model_type:`chat`,performance:1,speed:2,tagline:`Legacy GPT model for cheaper chat and non-chat tasks`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[3,4]},{label:`Batch`,values:[1.5,2]}]}],input:3,output:4,batch_input:1.5,batch_output:2},endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-3.5-turbo-16k-0613`]},{id:`gpt-3.5-turbo-instruct`,name:`gpt-3.5-turbo-instruct`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-3.5`,description:`Similar capabilities as GPT-3 era models. Compatible with legacy Completions endpoint and not Chat Completions.`,status:`deprecated`,knowledge_cutoff:`2021-08`,context_window:4096,max_output_tokens:4096,model_type:`chat`,performance:1,speed:2,tagline:`An older model only compatible with the legacy Completions endpoint`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[1.5,2]}]}],input:1.5,output:2},endpoints:[`chat_completions`,`responses`],snapshots:[`gpt-3.5-turbo-instruct`]},{id:`gpt-3.5-turbo`,name:`GPT-3.5 Turbo`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-3.5`,description:`GPT-3.5 Turbo models can understand and generate natural language or code and have been optimized for chat using the Chat Completions API but work well for non-chat tasks as well. As of July 2024, use gpt-4o-mini in place of GPT-3.5 Turbo, as it is cheaper, more capable, multimodal, and just as fast. GPT-3.5 Turbo is still available for use in the API.`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,model_type:`chat`,performance:1,speed:2,tagline:`Legacy GPT model for cheaper chat and non-chat tasks`,successor:`gpt-4o-mini`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,3,null,6]},{label:`Batch`,values:[null,1.5,null,3]}]},{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[.5,1.5]},{label:`Batch`,values:[.25,.75]}]}],input:.5,output:1.5,batch_input:.25,batch_output:.75},endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`],snapshots:[`gpt-3.5-turbo-0125`,`gpt-3.5-turbo-1106`,`gpt-3.5-turbo-instruct`]},{id:`gpt-4-0613`,name:`gpt-4-0613`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4`,status:`active`,knowledge_cutoff:`2023-11`,context_window:8192,max_output_tokens:8192,performance:2,speed:3,capabilities:{streaming:!0,fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`]},{id:`gpt-4-1106-vision-preview`,name:`gpt-4-1106-vision-preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4`,status:`active`,knowledge_cutoff:`2023-11`,context_window:128e3,max_output_tokens:4096,performance:2,speed:3,capabilities:{streaming:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`]},{id:`gpt-4-32k`,name:`gpt-4-32k`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4`,status:`active`,pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[60,120]},{label:`Batch`,values:[30,60]}]}],input:60,output:120,batch_input:30,batch_output:60}},{id:`gpt-4-turbo-2024-04-09`,name:`gpt-4-turbo-2024-04-09`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4-turbo`,status:`active`,knowledge_cutoff:`2023-11`,context_window:128e3,max_output_tokens:4096,alias:`gpt-4-turbo`,performance:2,speed:3,capabilities:{streaming:!0,vision:!0,tool_call:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],description:`GPT-4 Turbo is the next generation of GPT-4, an older high-intelligence GPT model. It was designed to be a cheaper, better version of GPT-4. Today, we recommend using a newer model like GPT-4o.`,model_type:`chat`,tagline:`GPT-4 Turbo is the next generation of GPT-4, an older high-intelligence GPT model.`,pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[10,30]},{label:`Batch`,values:[5,15]}]}],input:10,output:30,batch_input:5,batch_output:15}},{id:`gpt-4-turbo-preview`,name:`GPT-4 Turbo Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4-turbo`,description:`This is a research preview of the GPT-4 Turbo model, an older high-intelligence GPT model.`,status:`deprecated`,model_type:`chat`,tagline:`An older fast GPT model`,successor:`gpt-4o`,snapshots:[`gpt-4-0125-preview`,`gpt-4-1106-vision-preview`]},{id:`gpt-4-turbo`,name:`GPT-4 Turbo`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4-turbo`,status:`active`,knowledge_cutoff:`2023-11`,context_window:128e3,max_output_tokens:4096,performance:2,speed:3,capabilities:{streaming:!0,vision:!0,tool_call:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`gpt-4-turbo-2024-04-09`],description:`GPT-4 Turbo is the next generation of GPT-4, an older high-intelligence GPT model. It was designed to be a cheaper, better version of GPT-4. Today, we recommend using a newer model like GPT-4o.`,model_type:`chat`,tagline:`An older high-intelligence GPT model`,successor:`gpt-4o`,pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[10,30]},{label:`Batch`,values:[5,15]}]}],input:10,output:30,batch_input:5,batch_output:15}},{id:`gpt-4.1-2025-04-14`,name:`gpt-4.1-2025-04-14`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4.1`,description:`GPT-4.1 excels at instruction following and tool calling, with broad`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,alias:`gpt-4.1`,performance:4,speed:3,tagline:`GPT-4.1 excels at instruction following and tool calling, with broad`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3,.75,12]},{label:`Batch`,values:[null,1.5,null,6]}]}],input:2,output:8,cached_input:.5,batch_input:1,batch_output:4},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`]},{id:`gpt-4.1-mini-2025-04-14`,name:`gpt-4.1-mini-2025-04-14`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4.1`,description:`GPT-4.1 mini excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,alias:`gpt-4.1-mini`,performance:3,speed:4,tagline:`GPT-4.1 mini excels at instruction following and tool calling.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.8,.2,3.2]},{label:`Batch`,values:[null,.4,null,1.6]}]}],input:.4,output:1.6,cached_input:.1,batch_input:.2,batch_output:.8},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`]},{id:`gpt-4.1-mini`,name:`GPT-4.1 mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4.1`,description:`GPT-4.1 mini excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,performance:3,speed:4,tagline:`Smaller, faster version of GPT-4.1`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.4,output:1.6,cached_input:.1,batch_input:.2,batch_output:.8,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.8,.2,3.2]},{label:`Batch`,values:[null,.4,null,1.6]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4.1-mini-2025-04-14`]},{id:`gpt-4.1-nano-2025-04-14`,name:`gpt-4.1-nano-2025-04-14`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4.1`,description:`GPT-4.1 nano excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,alias:`gpt-4.1-nano`,performance:2,speed:5,tagline:`GPT-4.1 nano excels at instruction following and tool calling.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.5,.2,.05,.8]},{label:`Batch`,values:[null,.1,null,.4]}]}],input:.1,output:.4,cached_input:.025,batch_input:.05,batch_output:.2},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`]},{id:`gpt-4.1-nano`,name:`GPT-4.1 nano`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4.1`,description:`GPT-4.1 nano excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,performance:2,speed:5,tagline:`Fastest, most cost-efficient version of GPT-4.1`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.1,output:.4,cached_input:.025,batch_input:.05,batch_output:.2,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.5,.2,.05,.8]},{label:`Batch`,values:[null,.1,null,.4]}]}]},tools:[`function_calling`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4.1-nano-2025-04-14`]},{id:`gpt-4.1`,name:`GPT-4.1`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4.1`,description:`GPT-4.1 excels at instruction following and tool calling, with broad`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,performance:4,speed:3,tagline:`Smartest non-reasoning model`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3,.75,12]},{label:`Batch`,values:[null,1.5,null,6]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4.1-2025-04-14`]},{id:`gpt-4.5-preview-2025-02-27`,name:`gpt-4.5-preview-2025-02-27`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`Deprecated - a research preview of GPT-4.5. We recommend using gpt-4.1 or o3`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4.5-preview`,performance:4,speed:3,tagline:`Deprecated - a research preview of GPT-4.5.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:75,output:150,cached_input:37.5,batch_input:37.5,batch_output:75,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[75,37.5,150]},{label:`Batch`,values:[37.5,null,75]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`]},{id:`gpt-4.5-preview`,name:`GPT-4.5 Preview (Deprecated)`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`Deprecated - a research preview of GPT-4.5. We recommend using gpt-4.1 or o3`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:3,tagline:`Deprecated large model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:75,output:150,cached_input:37.5,batch_input:37.5,batch_output:75,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[75,37.5,150]},{label:`Batch`,values:[37.5,null,75]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`gpt-4.5-preview-2025-02-27`]},{id:`gpt-4`,name:`GPT-4`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4`,description:`GPT-4 is an older version of a high-intelligence GPT model, usable in Chat Completions.`,status:`active`,knowledge_cutoff:`2023-11`,context_window:8192,max_output_tokens:8192,model_type:`chat`,performance:2,speed:3,tagline:`An older high-intelligence GPT model`,successor:`gpt-4o`,capabilities:{streaming:!0,fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[30,60]},{label:`Batch`,values:[15,30]}]}],input:30,output:60,batch_input:15,batch_output:30},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4-0613`,`gpt-4-0314`]},{id:`gpt-4o-2024-05-13`,name:`gpt-4o-2024-05-13`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:4096,model_type:`chat`,alias:`gpt-4o`,performance:3,speed:3,tagline:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,1.25,10]},{label:`Batch`,values:[1.25,null,5]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`]},{id:`gpt-4o-2024-08-06`,name:`gpt-4o-2024-08-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o`,performance:3,speed:3,tagline:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3.75,1.875,15]},{label:`Batch`,values:[null,1.875,null,7.5]}]}],input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`]},{id:`gpt-4o-2024-11-20`,name:`gpt-4o-2024-11-20`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o`,performance:3,speed:3,tagline:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3.75,1.875,15]},{label:`Batch`,values:[null,1.875,null,7.5]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`]},{id:`gpt-4o-audio-preview-2024-10-01`,name:`gpt-4o-audio-preview-2024-10-01`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-audio-preview`,performance:3,speed:3,tagline:`This is a preview release of the GPT-4o Audio models.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`]},{id:`gpt-4o-audio-preview-2024-12-17`,name:`gpt-4o-audio-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-audio-preview`,performance:3,speed:3,tagline:`This is a preview release of the GPT-4o Audio models.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`]},{id:`gpt-4o-audio-preview-2025-06-03`,name:`gpt-4o-audio-preview-2025-06-03`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-audio-preview`,performance:3,speed:3,tagline:`This is a preview release of the GPT-4o Audio models.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`]},{id:`gpt-4o-audio-preview`,name:`GPT-4o Audio`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:3,speed:3,tagline:`GPT-4o models capable of audio inputs and outputs`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-4o-audio-preview-2025-06-03`,`gpt-4o-audio-preview-2024-12-17`,`gpt-4o-audio-preview-2024-10-01`]},{id:`gpt-4o-mini-2024-07-18`,name:`gpt-4o-mini-2024-07-18`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o mini (“o” for “omni”) is a fast, affordable small model for focused tasks.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-mini`,performance:2,speed:4,tagline:`GPT-4o mini (“o” for “omni”) is a fast, affordable small model for focused tasks.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,.3,.15,1.2]},{label:`Batch`,values:[null,.15,null,.6]}]}],input:.15,output:.6,cached_input:.075,batch_input:.075,batch_output:.3},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`]},{id:`gpt-4o-mini-audio-preview-2024-12-17`,name:`gpt-4o-mini-audio-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the smaller GPT-4o Audio mini model. It's designed to input audio or create audio outputs via the REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-mini-audio-preview`,performance:2,speed:4,tagline:`This is a preview release of the smaller GPT-4o Audio mini model.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},endpoints:[`chat_completions`]},{id:`gpt-4o-mini-audio-preview`,name:`GPT-4o mini Audio`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the smaller GPT-4o Audio mini model. It's designed to input audio or create audio outputs via the REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:2,speed:4,tagline:`Smaller model capable of audio inputs and outputs`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},tools:[`web_search`,`file_search`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`],snapshots:[`gpt-4o-mini-audio-preview-2024-12-17`]},{id:`gpt-4o-mini-realtime-preview-2024-12-17`,name:`gpt-4o-mini-realtime-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o-mini Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,alias:`gpt-4o-mini-realtime-preview`,performance:2,speed:5,tagline:`This is a preview release of the GPT-4o-mini Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.3,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.3,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]}]},endpoints:[`realtime`]},{id:`gpt-4o-mini-realtime-preview`,name:`GPT-4o mini Realtime`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o-mini Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,performance:2,speed:5,tagline:`Smaller realtime model for text and audio inputs and outputs`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.3,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.3,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]}]},endpoints:[`realtime`],snapshots:[`gpt-4o-mini-realtime-preview-2024-12-17`]},{id:`gpt-4o-mini-search-preview-2025-03-11`,name:`gpt-4o-mini-search-preview-2025-03-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o mini Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,alias:`gpt-4o-mini-search-preview`,performance:2,speed:4,tagline:`GPT-4o mini Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API.`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]}]},endpoints:[`chat_completions`]},{id:`gpt-4o-mini-search-preview`,name:`GPT-4o mini Search Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o mini Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,performance:2,speed:4,tagline:`Fast, affordable small model for web search`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-4o-mini-search-preview-2025-03-11`]},{id:`gpt-4o-mini-transcribe-2025-03-20`,name:`gpt-4o-mini-transcribe-2025-03-20`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,alias:`gpt-4o-mini-transcribe`,performance:3,speed:4,modalities:{input:[`audio`,`text`],output:[`text`]},endpoints:[`transcription`,`realtime`],description:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,tagline:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[1.25,5,.003]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[3,null,.003]}]}],input:1.25,output:5}},{id:`gpt-4o-mini-transcribe-2025-12-15`,name:`gpt-4o-mini-transcribe-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,alias:`gpt-4o-mini-transcribe`,performance:3,speed:4,modalities:{input:[`audio`,`text`],output:[`text`]},endpoints:[`transcription`,`realtime`],description:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,tagline:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[1.25,5,.003]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[3,null,.003]}]}],input:1.25,output:5}},{id:`gpt-4o-mini-transcribe`,name:`GPT-4o mini Transcribe`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,performance:3,speed:4,modalities:{input:[`audio`,`text`],output:[`text`]},endpoints:[`transcription`,`realtime`],snapshots:[`gpt-4o-mini-transcribe-2025-03-20`,`gpt-4o-mini-transcribe-2025-12-15`],description:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,tagline:`Speech-to-text model powered by GPT-4o mini`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[1.25,5,.003]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[3,null,.003]}]}],input:1.25,output:5}},{id:`gpt-4o-mini-tts-2025-03-20`,name:`gpt-4o-mini-tts-2025-03-20`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,status:`active`,model_type:`tts`,alias:`gpt-4o-mini-tts`,performance:4,speed:4,modalities:{input:[`text`],output:[`audio`]},endpoints:[`speech_generation`],description:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model. Use it to convert text to natural sounding spoken text. The maximum number of input tokens is 2000.`,tagline:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[.6,null,.015]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[null,12,.015]}]}],input:.6}},{id:`gpt-4o-mini-tts-2025-12-15`,name:`gpt-4o-mini-tts-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,status:`active`,model_type:`tts`,alias:`gpt-4o-mini-tts`,performance:4,speed:4,modalities:{input:[`text`],output:[`audio`]},endpoints:[`speech_generation`],description:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model. Use it to convert text to natural sounding spoken text. The maximum number of input tokens is 2000.`,tagline:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[.6,null,.015]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[null,12,.015]}]}],input:.6}},{id:`gpt-4o-mini-tts`,name:`GPT-4o mini TTS`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,status:`active`,model_type:`tts`,performance:4,speed:4,modalities:{input:[`text`],output:[`audio`]},endpoints:[`speech_generation`],snapshots:[`gpt-4o-mini-tts-2025-03-20`,`gpt-4o-mini-tts-2025-12-15`],description:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model. Use it to convert text to natural sounding spoken text. The maximum number of input tokens is 2000.`,tagline:`Text-to-speech model powered by GPT-4o mini`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[.6,null,.015]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[null,12,.015]}]}],input:.6}},{id:`gpt-4o-mini`,name:`GPT-4o mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o mini (“o” for “omni”) is a fast, affordable small model for focused tasks.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:2,speed:4,tagline:`Fast, affordable small model for focused tasks`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.15,output:.6,cached_input:.075,batch_input:.075,batch_output:.3,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,.3,.15,1.2]},{label:`Batch`,values:[null,.15,null,.6]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4o-mini-2024-07-18`]},{id:`gpt-4o-realtime-preview-2024-10-01`,name:`gpt-4o-realtime-preview-2024-10-01`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,alias:`gpt-4o-realtime-preview`,performance:2,speed:4,tagline:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`]},{id:`gpt-4o-realtime-preview-2024-12-17`,name:`gpt-4o-realtime-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,alias:`gpt-4o-realtime-preview`,performance:3,speed:4,tagline:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`]},{id:`gpt-4o-realtime-preview-2025-06-03`,name:`gpt-4o-realtime-preview-2025-06-03`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-4o-realtime-preview`,performance:3,speed:4,tagline:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`]},{id:`gpt-4o-realtime-preview`,name:`GPT-4o Realtime`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,performance:3,speed:4,tagline:`Model capable of realtime text and audio inputs and outputs`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`],snapshots:[`gpt-4o-realtime-preview-2025-06-03`,`gpt-4o-realtime-preview-2024-12-17`,`gpt-4o-realtime-preview-2024-10-01`]},{id:`gpt-4o-search-preview-2025-03-11`,name:`gpt-4o-search-preview-2025-03-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,alias:`gpt-4o-search-preview`,performance:3,speed:3,tagline:`GPT-4o Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API.`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]}]},endpoints:[`chat_completions`]},{id:`gpt-4o-search-preview`,name:`GPT-4o Search Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,performance:3,speed:3,tagline:`GPT model for web search in Chat Completions`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-4o-search-preview-2025-03-11`]},{id:`gpt-4o-transcribe-diarize`,name:`GPT-4o Transcribe Diarize`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o Transcribe Diarize is an automatic speech recognition (ASR) model with built-in speaker diarization, meaning it associates audio segments with different speakers in a conversation. This model is only available in the Transcription API.`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,performance:4,speed:3,tagline:`Transcription model that identifies who's speaking when`,modalities:{input:[`text`,`audio`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[2.5,10,.006]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[6,null,.006]}]}],input:2.5,output:10},endpoints:[`transcription`],snapshots:[`gpt-4o-transcribe-diarize`]},{id:`gpt-4o-transcribe`,name:`GPT-4o Transcribe`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o Transcribe is a speech-to-text model that uses GPT-4o to transcribe audio.`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,performance:4,speed:3,tagline:`Speech-to-text model powered by GPT-4o`,modalities:{input:[`audio`,`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[2.5,10,.006]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[6,null,.006]}]}],input:2.5,output:10},endpoints:[`transcription`,`realtime`],snapshots:[`gpt-4o-transcribe`]},{id:`gpt-4o`,name:`GPT-4o`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:3,speed:3,tagline:`Fast, intelligent, flexible GPT model`,capabilities:{streaming:!0,vision:!0,tool_call:!0,fine_tuning:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3.75,1.875,15]},{label:`Batch`,values:[null,1.875,null,7.5]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`gpt-4o-2024-11-20`,`gpt-4o-2024-08-06`,`gpt-4o-2024-05-13`]},{id:`gpt-5-2025-08-07`,name:`gpt-5-2025-08-07`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5 is our previous model for coding, reasoning, and agentic tasks across domains.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5`,performance:4,reasoning:4,speed:3,tagline:`GPT-5 is our previous model for coding, reasoning, and agentic tasks across domains.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},endpoints:[`chat_completions`,`responses`,`batch`]},{id:`gpt-5-chat-latest`,name:`GPT-5 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5 Chat points to the GPT-5 snapshot previously used in ChatGPT.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5 model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`],snapshots:[`gpt-5-chat-latest`]},{id:`gpt-5-codex`,name:`GPT-5-Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5-Codex is a version of GPT-5 optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`A version of GPT-5 optimized for agentic coding in Codex`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5-codex`]},{id:`gpt-5-latest`,name:`gpt-5-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,status:`active`,pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]}},{id:`gpt-5-mini-2025-08-07`,name:`gpt-5-mini-2025-08-07`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5 mini is a faster, more cost-efficient version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5-mini`,performance:3,reasoning:3,speed:4,tagline:`GPT-5 mini is a faster, more cost-efficient version of GPT-5.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.25,output:2,cached_input:.025,batch_input:.125,batch_output:1,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.25,.025,2]},{label:`Batch`,values:[.125,.0125,1]},{label:`Flex`,values:[.125,.0125,1]},{label:`Priority`,values:[.45,.05,3.6]}]}]},endpoints:[`chat_completions`,`responses`,`batch`]},{id:`gpt-5-mini`,name:`GPT-5 mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5 mini is a faster, more cost-efficient version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:3,reasoning:3,speed:4,tagline:`Near-frontier intelligence for cost sensitive, low latency, high volume workloads`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.25,output:2,cached_input:.025,batch_input:.125,batch_output:1,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.25,.025,2]},{label:`Batch`,values:[.125,.0125,1]},{label:`Flex`,values:[.125,.0125,1]},{label:`Priority`,values:[.45,.05,3.6]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-5-mini-2025-08-07`]},{id:`gpt-5-nano-2025-08-07`,name:`gpt-5-nano-2025-08-07`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5 Nano is our fastest, cheapest version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5-nano`,performance:2,reasoning:2,speed:5,tagline:`GPT-5 Nano is our fastest, cheapest version of GPT-5.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.05,output:.4,cached_input:.005,batch_input:.025,batch_output:.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.05,.005,.4]},{label:`Batch`,values:[.025,.0025,.2]},{label:`Flex`,values:[.025,.0025,.2]}]}]},endpoints:[`chat_completions`,`responses`,`batch`]},{id:`gpt-5-nano`,name:`GPT-5 nano`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5 Nano is our fastest, cheapest version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:2,reasoning:2,speed:5,tagline:`Fastest, most cost-efficient version of GPT-5`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.05,output:.4,cached_input:.005,batch_input:.025,batch_output:.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.05,.005,.4]},{label:`Batch`,values:[.025,.0025,.2]},{label:`Flex`,values:[.025,.0025,.2]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-5-nano-2025-08-07`]},{id:`gpt-5-pro-2025-10-06`,name:`gpt-5-pro-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:272e3,reasoning_tokens:!0,alias:`gpt-5-pro`,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:120,batch_input:7.5,batch_output:60,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,null,120]},{label:`Batch`,values:[7.5,null,60]}]}]},endpoints:[`responses`,`batch`]},{id:`gpt-5-pro`,name:`gpt-5-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:272e3,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:120,batch_input:7.5,batch_output:60,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,null,120]},{label:`Batch`,values:[7.5,null,60]}]}]},endpoints:[`responses`,`batch`],snapshots:[`gpt-5-pro-2025-10-06`]},{id:`gpt-5.1-2025-11-13`,name:`gpt-5.1-2025-11-13`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.1`,description:`GPT-5.1 is our flagship model for coding and agentic tasks with configurable reasoning and non-reasoning effort.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5.1`,performance:4,reasoning:4,speed:3,tagline:`GPT-5.1 is our flagship model for coding and agentic tasks with configurable reasoning and non-reasoning effort.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},endpoints:[`responses`,`chat_completions`,`batch`]},{id:`gpt-5.1-chat-latest`,name:`GPT-5.1 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.1`,description:`GPT-5.1 Chat points to the GPT-5.1 snapshot currently used in ChatGPT.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5.1 model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`responses`,`chat_completions`],snapshots:[`gpt-5.1-chat-latest`]},{id:`gpt-5.1-codex-max`,name:`GPT-5.1-Codex-Max`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.1`,description:`GPT‑5.1-Codex-Max is purpose-built for agentic coding.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:4,tagline:`A version of GPT-5.1-codex optimized for long running tasks.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5.1-codex-max`]},{id:`gpt-5.1-codex-mini`,name:`GPT-5.1 Codex mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.1`,description:`GPT-5.1 Codex mini is a smaller, more cost-effective, less-capable version of GPT-5.1-Codex.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Smaller, more cost-effective, less-capable version of GPT-5.1-Codex`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.25,output:2,cached_input:.025,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.25,.025,2]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5.1-codex-mini`]},{id:`gpt-5.1-codex`,name:`GPT-5.1 Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.1`,description:`GPT-5.1-Codex is a version of GPT-5 optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`A version of GPT-5.1 optimized for agentic coding in Codex.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5.1-codex`]},{id:`gpt-5.1-latest`,name:`gpt-5.1-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.1`,status:`active`,pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]}},{id:`gpt-5.1`,name:`GPT-5.1`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.1`,description:`GPT-5.1 is our flagship model for coding and agentic tasks with configurable reasoning and non-reasoning effort.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`The best model for coding and agentic tasks with configurable reasoning effort`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`apply_patch`,`mcp`],endpoints:[`responses`,`chat_completions`,`batch`],snapshots:[`gpt-5.1-2025-11-13`]},{id:`gpt-5.2-2025-12-11`,name:`gpt-5.2-2025-12-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.2`,description:`GPT-5.2 is our previous frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5.2`,performance:5,reasoning:5,speed:3,tagline:`GPT-5.2 is our previous frontier model for complex professional work.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,batch_input:.875,batch_output:7,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]},{label:`Batch`,values:[.875,.0875,7]},{label:`Flex`,values:[.875,.0875,7]},{label:`Priority`,values:[3.5,.35,28]}]}]},endpoints:[`responses`,`chat_completions`,`batch`]},{id:`gpt-5.2-chat-latest`,name:`GPT-5.2 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.2`,description:`GPT-5.2 Chat points to the GPT-5.2 snapshot currently used in ChatGPT.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5.2 model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`responses`,`chat_completions`],snapshots:[`gpt-5.2-chat-latest`]},{id:`gpt-5.2-codex`,name:`GPT-5.2-Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.2`,description:"GPT-5.2-Codex is an upgraded version of GPT-5.2 optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.\\nGPT-5.2-Codex supports `low`, `medium`, `high`, and `xhigh` reasoning effort settings.\\nIf you want to learn more about prompting GPT-5.2-Codex, refer to our [dedicated guide](/cookbook/examples/gpt-5/codex_prompting_guide).\\n",status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Our most intelligent coding model optimized for long-horizon, agentic coding tasks.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`hosted_shell`,`skills`],endpoints:[`responses`],snapshots:[`gpt-5.2-codex`]},{id:`gpt-5.2-pro-2025-12-11`,name:`gpt-5.2-pro-2025-12-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.2`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,reasoning_tokens:!0,alias:`gpt-5.2-pro`,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:21,output:168,batch_input:10.5,batch_output:84,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[21,null,168]},{label:`Batch`,values:[10.5,null,84]},{label:`Flex`,values:[10.5,null,84]}]}]},endpoints:[`responses`]},{id:`gpt-5.2-pro`,name:`gpt-5.2-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.2`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:21,output:168,batch_input:10.5,batch_output:84,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[21,null,168]},{label:`Batch`,values:[10.5,null,84]},{label:`Flex`,values:[10.5,null,84]}]}]},endpoints:[`responses`],snapshots:[`gpt-5.2-pro-2025-12-11`]},{id:`gpt-5.2`,name:`GPT-5.2`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.2`,description:`GPT-5.2 is our previous frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:5,reasoning:5,speed:3,tagline:`Previous frontier model for professional work with configurable reasoning effort`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,batch_input:.875,batch_output:7,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]},{label:`Batch`,values:[.875,.0875,7]},{label:`Flex`,values:[.875,.0875,7]},{label:`Priority`,values:[3.5,.35,28]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`hosted_shell`,`apply_patch`,`skills`,`mcp`],endpoints:[`responses`,`chat_completions`,`batch`],snapshots:[`gpt-5.2-2025-12-11`]},{id:`gpt-5.3-chat-latest`,name:`GPT-5.3 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.3`,description:`GPT-5.3 Chat points to the GPT-5.3 Instant snapshot currently used in ChatGPT.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5.3 Instant model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`responses`,`chat_completions`],snapshots:[`gpt-5.3-chat-latest`]},{id:`gpt-5.3-codex`,name:`GPT-5.3-Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.3`,description:"GPT-5.3-Codex is optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.\\nGPT-5.3-Codex supports `low`, `medium`, `high`, and `xhigh` reasoning effort settings.\\nIf you want to learn more about prompting GPT-5.3-Codex, refer to our [dedicated guide](/cookbook/examples/gpt-5/codex_prompting_guide).\\n",status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`The most capable agentic coding model to date.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`hosted_shell`,`skills`],endpoints:[`responses`],snapshots:[`gpt-5.3-codex`]},{id:`gpt-5.4-2026-03-05`,name:`gpt-5.4-2026-03-05`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,description:`GPT-5.4 is our frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5.4`,performance:5,reasoning:5,speed:3,tagline:`GPT-5.4 is our frontier model for complex professional work.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:15,cached_input:.25,batch_input:1.25,batch_output:7.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,.25,15]},{label:`Batch`,values:[1.25,.13,7.5]},{label:`Flex`,values:[1.25,.13,7.5]},{label:`Priority`,values:[5,.5,30]}]}]},endpoints:[`responses`,`chat_completions`,`batch`]},{id:`gpt-5.4-mini-2026-03-17`,name:`gpt-5.4-mini-2026-03-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,description:`GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5.4-mini`,performance:4,reasoning:4,speed:4,tagline:`GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.75,output:4.5,cached_input:.075,batch_input:.375,batch_output:2.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.75,.075,4.5]},{label:`Batch`,values:[.375,.0375,2.25]},{label:`Flex`,values:[.375,.0375,2.25]}]}]},endpoints:[`responses`,`chat_completions`,`batch`]},{id:`gpt-5.4-mini`,name:`GPT-5.4 mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,description:`GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:4,tagline:`Our strongest mini model yet for coding, computer use, and subagents`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.75,output:4.5,cached_input:.075,batch_input:.375,batch_output:2.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.75,.075,4.5]},{label:`Batch`,values:[.375,.0375,2.25]},{label:`Flex`,values:[.375,.0375,2.25]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`tool_search`,`image_generation`,`code_interpreter`,`hosted_shell`,`apply_patch`,`skills`,`computer_use`,`mcp`],endpoints:[`responses`,`chat_completions`,`batch`],pricing_notes:[`Regional processing (data residency) endpoints are charged a 10% uplift for GPT-5.4 mini.`],snapshots:[`gpt-5.4-mini-2026-03-17`]},{id:`gpt-5.4-nano-2026-03-17`,name:`gpt-5.4-nano-2026-03-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,description:`GPT-5.4 nano is designed for tasks where speed and cost matter most like`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,alias:`gpt-5.4-nano`,performance:3,reasoning:3,speed:4,tagline:`GPT-5.4 nano is designed for tasks where speed and cost matter most like`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.2,output:1.25,cached_input:.02,batch_input:.1,batch_output:.625,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.2,.02,1.25]},{label:`Batch`,values:[.1,.01,.625]},{label:`Flex`,values:[.1,.01,.625]}]}]},endpoints:[`responses`,`chat_completions`,`batch`]},{id:`gpt-5.4-nano`,name:`GPT-5.4 nano`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,description:`GPT-5.4 nano is designed for tasks where speed and cost matter most like`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:3,reasoning:3,speed:4,tagline:`Our cheapest GPT-5.4-class model for simple high-volume tasks`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.2,output:1.25,cached_input:.02,batch_input:.1,batch_output:.625,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.2,.02,1.25]},{label:`Batch`,values:[.1,.01,.625]},{label:`Flex`,values:[.1,.01,.625]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`hosted_shell`,`apply_patch`,`skills`,`mcp`],endpoints:[`responses`,`chat_completions`,`batch`],pricing_notes:[`Regional processing (data residency) endpoints are charged a 10% uplift for GPT-5.4 nano.`],snapshots:[`gpt-5.4-nano-2026-03-17`]},{id:`gpt-5.4-pro-2026-03-05`,name:`gpt-5.4-pro-2026-03-05`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,reasoning_tokens:!0,alias:`gpt-5.4-pro`,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:30,output:180,batch_input:15,batch_output:90,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[30,null,180]},{label:`Batch`,values:[15,null,90]},{label:`Flex`,values:[15,null,90]}]}]},endpoints:[`responses`]},{id:`gpt-5.4-pro`,name:`gpt-5.4-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:30,output:180,batch_input:15,batch_output:90,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[30,null,180]},{label:`Batch`,values:[15,null,90]},{label:`Flex`,values:[15,null,90]}]}]},endpoints:[`responses`],snapshots:[`gpt-5.4-pro-2026-03-05`]},{id:`gpt-5.4`,name:`GPT-5.4`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5.4`,description:`GPT-5.4 is our frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:5,reasoning:5,speed:3,tagline:`Best intelligence at scale for agentic, coding, and professional workflows`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:15,cached_input:.25,batch_input:1.25,batch_output:7.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,.25,15]},{label:`Batch`,values:[1.25,.13,7.5]},{label:`Flex`,values:[1.25,.13,7.5]},{label:`Priority`,values:[5,.5,30]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`tool_search`,`image_generation`,`code_interpreter`,`hosted_shell`,`apply_patch`,`skills`,`computer_use`,`mcp`],endpoints:[`responses`,`chat_completions`,`batch`],pricing_notes:[`For models with a 1.05M context window (GPT-5.4 and GPT-5.4 pro), prompts with >272K input tokens are priced at 2x input and 1.5x output for the full session for standard, batch, and flex.`,`Regional processing (data residency) endpoints are charged a 10% uplift for GPT-5.4 and GPT-5.4 pro.`],snapshots:[`gpt-5.4-2026-03-05`]},{id:`gpt-5`,name:`GPT-5`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-5`,description:`GPT-5 is our previous model for coding, reasoning, and agentic tasks across domains.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Previous intelligent reasoning model for coding and agentic tasks with configurable reasoning effort`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-5-2025-08-07`]},{id:`gpt-audio-1.5`,name:`gpt-audio-1.5`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-audio`,description:`The gpt-audio model is our first generally available audio model. It accepts`,status:`active`,knowledge_cutoff:`2024-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:3,tagline:`The best voice model for audio in, audio out with Chat Completions.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,null,64]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-audio-1.5`]},{id:`gpt-audio-2025-08-28`,name:`gpt-audio-2025-08-28`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-audio`,description:`The gpt-audio model is our first generally available audio model. It accepts`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-audio`,performance:4,speed:3,tagline:`The gpt-audio model is our first generally available audio model.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,null,64]}]}]},endpoints:[`chat_completions`]},{id:`gpt-audio-mini-2025-10-06`,name:`gpt-audio-mini-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-audio`,description:`A cost-efficient version of GPT Audio. It accepts audio inputs and outputs, and can be used in the Chat Completions REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-audio-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Audio.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,null,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},endpoints:[`chat_completions`]},{id:`gpt-audio-mini-2025-12-15`,name:`gpt-audio-mini-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-audio`,description:`A cost-efficient version of GPT Audio. It accepts audio inputs and outputs, and can be used in the Chat Completions REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-audio-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Audio.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,null,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},endpoints:[`chat_completions`]},{id:`gpt-audio-mini`,name:`gpt-audio-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-audio`,description:`A cost-efficient version of GPT Audio. It accepts audio inputs and outputs, and can be used in the Chat Completions REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Audio`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,null,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`],snapshots:[`gpt-audio-mini-2025-10-06`,`gpt-audio-mini-2025-12-15`]},{id:`gpt-audio`,name:`gpt-audio`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-audio`,description:`The gpt-audio model is our first generally available audio model. It accepts`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:3,tagline:`For audio inputs and outputs with Chat Completions API`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,null,64]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-audio-2025-08-28`]},{id:`gpt-image-1-mini`,name:`gpt-image-1-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-image-1`,description:`A cost-efficient version of GPT Image 1. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.`,status:`active`,model_type:`image`,performance:4,speed:1,tagline:`A cost-efficient version of GPT Image 1`,modalities:{input:[`text`,`image`],output:[`image`]},pricing:{input:2,cached_input:.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.2,null]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,.25,8]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1792`,`1792x1024`],rows:[{label:`Low`,values:[null,.005,.006,.006]},{label:`Medium`,values:[null,.011,.015,.015]},{label:`High`,values:[null,.036,.052,.052]}]}]},endpoints:[`batch`,`image_edit`,`image_generation`],snapshots:[`gpt-image-1-mini`]},{id:`gpt-image-1.5-2025-12-16`,name:`gpt-image-1.5-2025-12-16`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-image-1.5`,description:`GPT Image 1.5 is our latest image generation model, with better instruction following and adherence to prompts. Learn more in our [GPT Image 1.5 usage guide](/api/docs/guides/image-generation).`,status:`active`,model_type:`image`,alias:`gpt-image-1.5`,performance:5,speed:3,tagline:`GPT Image 1.5 is our latest image generation model, with better instruction following and adherence to prompts.`,modalities:{input:[`text`,`image`],output:[`image`,`text`]},pricing:{input:5,output:10,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,10]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,2,32]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.009,.013,.013]},{label:`Medium`,values:[null,.034,.05,.05]},{label:`High`,values:[null,.133,.2,.2]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`]},{id:`gpt-image-1.5`,name:`GPT Image 1.5`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-image-1.5`,description:`GPT Image 1.5 is our latest image generation model, with better instruction following and adherence to prompts. Learn more in our [GPT Image 1.5 usage guide](/api/docs/guides/image-generation).`,status:`active`,model_type:`image`,performance:5,speed:3,tagline:`State-of-the-art image generation model.`,modalities:{input:[`text`,`image`],output:[`image`,`text`]},pricing:{input:5,output:10,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,10]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,2,32]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.009,.013,.013]},{label:`Medium`,values:[null,.034,.05,.05]},{label:`High`,values:[null,.133,.2,.2]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`],snapshots:[`gpt-image-1.5-2025-12-16`]},{id:`gpt-image-1`,name:`GPT Image 1`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-image-1`,description:`GPT Image 1 is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.`,status:`active`,model_type:`image`,performance:4,speed:1,tagline:`Our previous image generation model`,modalities:{input:[`text`,`image`],output:[`image`]},pricing:{input:5,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,null]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,2.5,40]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.011,.016,.016]},{label:`Medium`,values:[null,.042,.063,.063]},{label:`High`,values:[null,.167,.25,.25]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`,`responses`],snapshots:[`gpt-image-1`]},{id:`gpt-oss-120b`,name:`gpt-oss-120b`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-oss`,description:`\\`,status:`active`,knowledge_cutoff:`2024-05`,context_window:131072,max_output_tokens:131072,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Most powerful open-weight model, fits into an H100 GPU`,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},tools:[`function_calling`,`code_interpreter`,`mcp`,`web_search`],endpoints:[`responses`,`batch`],snapshots:[`gpt-oss-120b`]},{id:`gpt-oss-20b`,name:`gpt-oss-20b`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-oss`,description:`\\`,status:`active`,knowledge_cutoff:`2024-05`,context_window:131072,max_output_tokens:131072,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Medium-sized open-weight model for low latency`,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},tools:[`function_calling`,`code_interpreter`,`mcp`,`web_search`],endpoints:[`responses`,`batch`],snapshots:[`gpt-oss-20b`]},{id:`gpt-realtime-1.5`,name:`gpt-realtime-1.5`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-realtime`,description:`GPT-Reatime-1.5 is our flagship audio model for voice agents & customer support.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:32e3,max_output_tokens:4096,performance:5,speed:4,tagline:`The best voice model for audio in, audio out.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`,`image`],output:[`text`,`audio`]},pricing:{input:4,output:16,cached_input:.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[4,.4,16]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,.4,64]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.5,null]}]}]},endpoints:[`realtime`],snapshots:[`gpt-realtime-1.5`]},{id:`gpt-realtime-2025-08-28`,name:`gpt-realtime-2025-08-28`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-realtime`,description:`This is our first general-availability realtime model, capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-realtime`,performance:5,speed:4,tagline:`This is our first general-availability realtime model, capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`,`image`],output:[`text`,`audio`]},pricing:{input:4,output:16,cached_input:.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[4,.4,16]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,.4,64]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.5,null]}]}]},endpoints:[`realtime`]},{id:`gpt-realtime-mini-2025-10-06`,name:`gpt-realtime-mini-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-realtime`,description:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-realtime-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`image`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.06,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.06,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.8,.08,null]}]}]},endpoints:[`realtime`]},{id:`gpt-realtime-mini-2025-12-15`,name:`gpt-realtime-mini-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-realtime`,description:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-realtime-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`image`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.06,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.06,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.8,.08,null]}]}]},endpoints:[`realtime`]},{id:`gpt-realtime-mini`,name:`gpt-realtime-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-realtime`,description:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,performance:4,speed:5,tagline:`A cost-efficient version of GPT Realtime`,capabilities:{tool_call:!0},modalities:{input:[`text`,`image`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.06,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.06,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.8,.08,null]}]}]},endpoints:[`realtime`],snapshots:[`gpt-realtime-mini-2025-10-06`,`gpt-realtime-mini-2025-12-15`]},{id:`gpt-realtime`,name:`gpt-realtime`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`gpt-realtime`,description:`This is our first general-availability realtime model, capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,performance:5,speed:4,tagline:`Model capable of realtime text and audio inputs and outputs`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`,`image`],output:[`text`,`audio`]},pricing:{input:4,output:16,cached_input:.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[4,.4,16]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,.4,64]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.5,null]}]}]},endpoints:[`realtime`],snapshots:[`gpt-realtime-2025-08-28`]},{id:`o1-2024-12-17`,name:`o1-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1`,performance:4,reasoning:4,speed:1,tagline:`The o1 series of models are trained with reinforcement learning to perform complex reasoning.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:60,cached_input:7.5,batch_input:7.5,batch_output:30,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,7.5,60]},{label:`Batch`,values:[7.5,null,30]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`]},{id:`o1-mini-2024-09-12`,name:`o1-mini-2024-09-12`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,description:`The o1 reasoning model is designed to solve hard problems across domains. o1-mini is a faster and more affordable reasoning model, but we recommend using the newer o3-mini model that features higher intelligence at the same latency and price as o1-mini.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:65536,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1-mini`,performance:3,reasoning:3,speed:2,tagline:`The o1 reasoning model is designed to solve hard problems across domains.`,capabilities:{streaming:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},endpoints:[`chat_completions`,`assistants`]},{id:`o1-mini`,name:`o1-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,description:`The o1 reasoning model is designed to solve hard problems across domains. o1-mini is a faster and more affordable reasoning model, but we recommend using the newer o3-mini model that features higher intelligence at the same latency and price as o1-mini.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:65536,model_type:`reasoning`,reasoning_tokens:!0,performance:3,reasoning:3,speed:2,tagline:`A small model alternative to o1`,successor:`o3-mini`,capabilities:{streaming:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},tools:[`file_search`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`assistants`],snapshots:[`o1-mini-2024-09-12`]},{id:`o1-preview-2024-09-12`,name:`o1-preview-2024-09-12`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:32768,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1-preview`,performance:3,reasoning:3,speed:1,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`assistants`],description:`Research preview of the o1 series of models, trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,tagline:`Research preview of the o1 series of models, trained with reinforcement learning to perform complex reasoning.`},{id:`o1-preview`,name:`o1 Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,status:`deprecated`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:32768,model_type:`reasoning`,reasoning_tokens:!0,performance:3,reasoning:3,speed:1,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`assistants`],snapshots:[`o1-preview-2024-09-12`],description:`Research preview of the o1 series of models, trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,tagline:`Preview of our first o-series reasoning model`,successor:`o1`},{id:`o1-pro-2025-03-19`,name:`o1-pro-2025-03-19`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1-pro`,performance:4,reasoning:4,speed:1,tagline:`The o1 series of models are trained with reinforcement learning to think`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:150,output:600,batch_input:75,batch_output:300,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[150,null,600]},{label:`Batch`,values:[75,null,300]}]}]},endpoints:[`responses`,`batch`]},{id:`o1-pro`,name:`o1-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:1,tagline:`Version of o1 with more compute for better responses`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:150,output:600,batch_input:75,batch_output:300,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[150,null,600]},{label:`Batch`,values:[75,null,300]}]}]},tools:[`function_calling`,`file_search`,`mcp`],endpoints:[`responses`,`batch`],snapshots:[`o1-pro-2025-03-19`]},{id:`o1`,name:`o1`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:1,tagline:`Previous full o-series reasoning model`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:60,cached_input:7.5,batch_input:7.5,batch_output:30,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,7.5,60]},{label:`Batch`,values:[7.5,null,30]}]}]},tools:[`function_calling`,`file_search`,`mcp`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`o1-2024-12-17`]},{id:`o3-2025-04-16`,name:`o3-2025-04-16`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`o3 is a well-rounded and powerful model across domains. It sets a new`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o3`,performance:5,reasoning:5,speed:1,tagline:`o3 is a well-rounded and powerful model across domains.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]},{label:`Flex`,values:[1,.25,4]}]}]},endpoints:[`chat_completions`,`responses`,`batch`]},{id:`o3-deep-research-2025-06-26`,name:`o3-deep-research-2025-06-26`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`o3-deep-research is our most advanced model for deep research,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o3-deep-research`,performance:5,reasoning:5,speed:1,tagline:`o3-deep-research is our most advanced model for deep research,`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:10,output:40,cached_input:2.5,batch_input:5,batch_output:20,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,2.5,40]},{label:`Batch`,values:[5,null,20]}]}]},endpoints:[`responses`,`batch`]},{id:`o3-deep-research`,name:`o3-deep-research`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`o3-deep-research is our most advanced model for deep research,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,tagline:`Our most powerful deep research model`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:10,output:40,cached_input:2.5,batch_input:5,batch_output:20,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,2.5,40]},{label:`Batch`,values:[5,null,20]}]}]},tools:[`web_search`,`code_interpreter`,`mcp`],endpoints:[`responses`,`batch`],snapshots:[`o3-deep-research-2025-06-26`]},{id:`o3-mini-2025-01-31`,name:`o3-mini-2025-01-31`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`o3-mini is our newest small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini. o3-mini supports key developer features, like Structured Outputs, function calling, and Batch API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o3-mini`,performance:4,reasoning:4,speed:3,tagline:`o3-mini is our newest small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini.`,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`]},{id:`o3-mini`,name:`o3-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`o3-mini is our newest small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini. o3-mini supports key developer features, like Structured Outputs, function calling, and Batch API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`A small model alternative to o3`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},tools:[`function_calling`,`file_search`,`code_interpreter`,`mcp`,`image_generation`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`o3-mini-2025-01-31`]},{id:`o3-pro-2025-06-10`,name:`o3-pro-2025-06-10`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`The o-series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o3-pro`,performance:5,reasoning:5,speed:1,tagline:`The o-series of models are trained with reinforcement learning to think`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:20,output:80,batch_input:10,batch_output:40,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[20,null,80]},{label:`Batch`,values:[10,null,40]}]}]},endpoints:[`responses`,`batch`]},{id:`o3-pro`,name:`o3-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`The o-series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,tagline:`Version of o3 with more compute for better responses`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:20,output:80,batch_input:10,batch_output:40,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[20,null,80]},{label:`Batch`,values:[10,null,40]}]}]},tools:[`function_calling`,`file_search`,`image_generation`,`mcp`,`web_search`],endpoints:[`responses`,`batch`],snapshots:[`o3-pro-2025-06-10`]},{id:`o3`,name:`o3`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o3`,description:`o3 is a well-rounded and powerful model across domains. It sets a new`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,tagline:`Reasoning model for complex tasks, succeeded by GPT-5`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]},{label:`Flex`,values:[1,.25,4]}]}]},tools:[`function_calling`,`file_search`,`image_generation`,`code_interpreter`,`mcp`,`web_search`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`o3-2025-04-16`]},{id:`o4-mini-2025-04-16`,name:`o4-mini-2025-04-16`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o4`,description:`o4-mini is our latest small o-series model. It's optimized for fast,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o4-mini`,performance:4,reasoning:4,speed:3,tagline:`o4-mini is our latest small o-series model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[null,4,1,16]},{label:`Batch`,values:[null,2,null,8]}]}],input:1.1,output:4.4,cached_input:.275,batch_input:.55,batch_output:2.2},endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`]},{id:`o4-mini-deep-research-2025-06-26`,name:`o4-mini-deep-research-2025-06-26`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o4`,description:`o4-mini-deep-research is our faster, more affordable deep`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o4-mini-deep-research`,performance:4,reasoning:4,speed:3,tagline:`o4-mini-deep-research is our faster, more affordable deep`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]}]}]},endpoints:[`responses`,`batch`]},{id:`o4-mini-deep-research`,name:`o4-mini-deep-research`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o4`,description:`o4-mini-deep-research is our faster, more affordable deep`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Faster, more affordable deep research model`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]}]}]},tools:[`web_search`,`code_interpreter`,`mcp`],endpoints:[`responses`,`batch`],snapshots:[`o4-mini-deep-research-2025-06-26`]},{id:`o4-mini`,name:`o4-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`o4`,description:`o4-mini is our latest small o-series model. It's optimized for fast,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Fast, cost-efficient reasoning model, succeeded by GPT-5 mini`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.275,batch_input:.55,batch_output:2.2,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[null,4,1,16]},{label:`Batch`,values:[null,2,null,8]}]}]},tools:[`function_calling`,`file_search`,`code_interpreter`,`mcp`,`web_search`],endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`],snapshots:[`o4-mini-2025-04-16`]},{id:`omni-moderation-2024-09-26`,name:`omni-moderation-2024-09-26`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,model_type:`moderation`,alias:`omni-moderation`,performance:3,speed:3,capabilities:{vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`batch`,`moderation`]},{id:`omni-moderation-latest`,name:`omni-moderation`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`Moderation models are free models designed to detect harmful content.`,status:`active`,model_type:`moderation`,performance:3,speed:3,tagline:`Identify potentially harmful content in text and images`,capabilities:{vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Moderation`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[null]}]}]},endpoints:[`batch`,`moderation`],snapshots:[`omni-moderation-2024-09-26`]},{id:`omni-moderation`,name:`omni-moderation`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,model_type:`moderation`,performance:3,speed:3,capabilities:{vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`batch`,`moderation`],snapshots:[`omni-moderation-2024-09-26`]},{id:`sora-2-2025-10-06`,name:`sora-2-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,model_type:`video`,alias:`sora-2`,performance:4,speed:2,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],description:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,tagline:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.1,null,null]}]}]}},{id:`sora-2-2025-12-08`,name:`sora-2-2025-12-08`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,model_type:`video`,alias:`sora-2`,performance:4,speed:2,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],description:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,tagline:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.1,null,null]}]}]}},{id:`sora-2-pro-2025-10-06`,name:`sora-2-pro-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,model_type:`video`,alias:`sora-2-pro`,performance:5,speed:1,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],description:`Sora 2 Pro is our state-of-the-art, most advanced media generation model, generating videos with synced audio.`,tagline:`Sora 2 Pro is our state-of-the-art, most advanced media generation model, generating videos with synced audio.`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.3,.5,.7]}]}]}},{id:`sora-2-pro`,name:`Sora 2 Pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,model_type:`video`,performance:5,speed:1,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],snapshots:[`sora-2-pro-2025-10-06`,`sora-2-pro`],description:`Sora 2 Pro is our state-of-the-art, most advanced media generation model, generating videos with synced audio.`,tagline:`Most advanced synced-audio video generation`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.3,.5,.7]}]}]}},{id:`sora-2`,name:`Sora 2`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,model_type:`video`,performance:4,speed:2,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],snapshots:[`sora-2-2025-10-06`,`sora-2-2025-12-08`,`sora-2`],description:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,tagline:`Flagship video generation with synced audio`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.1,null,null]}]}]}},{id:`text-embedding-3-large`,name:`text-embedding-3-large`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`text-embedding-3-large is our most capable embedding model for both english and non-english tasks.`,status:`active`,model_type:`embed`,performance:3,speed:2,tagline:`Most capable embedding model`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Embeddings`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.13]},{label:`Batch`,values:[.065]}]}]},endpoints:[`embeddings`,`batch`],snapshots:[`text-embedding-3-large`]},{id:`text-embedding-3-small`,name:`text-embedding-3-small`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`text-embedding-3-small is our improved, more performant version of our ada embedding model.`,status:`active`,model_type:`embed`,performance:2,speed:3,tagline:`Small embedding model`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Embeddings`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.02]},{label:`Batch`,values:[.01]}]}]},endpoints:[`embeddings`,`batch`],snapshots:[`text-embedding-3-small`]},{id:`text-embedding-ada-002`,name:`text-embedding-ada-002`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`text-embedding-ada-002 is our improved, more performant version of our ada embedding model.`,status:`active`,model_type:`embed`,performance:1,speed:2,tagline:`Older embedding model`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Embeddings`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.1]},{label:`Batch`,values:[.05]}]}]},endpoints:[`embeddings`,`batch`],snapshots:[`text-embedding-ada-002`]},{id:`text-moderation-007`,name:`text-moderation-007`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,status:`active`,knowledge_cutoff:`2021-08`,max_output_tokens:32768,model_type:`moderation`,performance:2,speed:3,modalities:{input:[`text`],output:[`text`]},endpoints:[`moderation`]},{id:`text-moderation-latest`,name:`text-moderation`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`Moderation models are free models designed to detect harmful content. This is our text only moderation model; we expect omni-moderation-* models to be the best default moving forward.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:32768,model_type:`moderation`,performance:2,speed:3,tagline:`Previous generation text-only moderation model`,successor:`omni-moderation`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Moderation`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[null]}]}]},endpoints:[`moderation`],snapshots:[`text-moderation-007`]},{id:`text-moderation-stable`,name:`text-moderation-stable`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`Moderation models are free models designed to detect harmful content. This is our text only moderation model; we expect omni-moderation-* models to be the best default moving forward.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:32768,model_type:`moderation`,performance:2,speed:3,tagline:`Previous generation text-only moderation model`,successor:`omni-moderation`,modalities:{input:[`text`],output:[`text`]},endpoints:[`moderation`],snapshots:[`text-moderation-007`]},{id:`tts-1-hd`,name:`TTS-1 HD`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`TTS is a model that converts text to natural sounding spoken text. The tts-1-hd model is optimized for high quality text-to-speech use cases. Use it with the Speech endpoint in the Audio API.`,status:`active`,model_type:`tts`,performance:3,speed:3,tagline:`Text-to-speech model optimized for quality`,modalities:{input:[`text`],output:[`audio`]},pricing:{tiers:[{label:`Pricing`,unit:`Per 1M tokens`,columns:[`Cost`],rows:[{label:`Standard`,values:[30]}]}]},endpoints:[`speech_generation`],snapshots:[`tts-1-hd`]},{id:`tts-1`,name:`TTS-1`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,description:`TTS is a model that converts text to natural sounding spoken text. The tts-1 model is optimized for realtime text-to-speech use cases. Use it with the Speech endpoint in the Audio API.`,status:`active`,model_type:`tts`,performance:2,speed:4,tagline:`Text-to-speech model optimized for speed`,modalities:{input:[`text`],output:[`audio`]},pricing:{tiers:[{label:`Pricing`,unit:`Per 1M tokens`,columns:[`Cost`],rows:[{label:`Standard`,values:[15]}]}]},endpoints:[`speech_generation`],snapshots:[`tts-1`]},{id:`whisper-1`,name:`Whisper`,created_by:`openai`,source:`official`,last_updated:`2026-03-21`,family:`whisper`,description:`Whisper is a general-purpose speech recognition model, trained on a large dataset of diverse audio. You can also use it as a multitask model to perform multilingual speech recognition as well as speech translation and language identification.`,status:`active`,model_type:`transcription`,performance:2,speed:3,tagline:`General-purpose speech recognition model`,modalities:{input:[`audio`],output:[`text`]},pricing:{tiers:[{label:`Pricing`,unit:`Per 1M tokens`,columns:[`Cost`],rows:[{label:`Standard`,values:[.006]}]}]},endpoints:[`transcription`,`translation`],snapshots:[`whisper-1`]}]},t=e.models.map(t=>({...t,provider:e.id}));exports.models=t,exports.provider=e;
+</svg>`,models:[{id:`babbage-002`,name:`babbage-002`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`GPT base models can understand and generate natural language or code but are not trained with instruction following. These models are made to be replacements for our original GPT-3 base models and use the legacy Completions API. Most customers should use GPT-3.5 or GPT-4.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:16384,model_type:`chat`,performance:1,speed:3,tagline:`Replacement for the GPT-3 ada and babbage base models`,successor:`gpt-4o`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.4,1.6,null,1.6]},{label:`Batch`,values:[null,.8,null,.8]}]},{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[.4,.4]},{label:`Batch`,values:[.2,.2]}]}],input:.4,output:.4,batch_input:.2,batch_output:.2},endpoints:[`completions`],snapshots:[`babbage-002`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/babbage-002`,open_weight:!1},{id:`chatgpt-4o-latest`,name:`ChatGPT-4o`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`chatgpt`,description:`ChatGPT-4o points to the GPT-4o snapshot currently used in ChatGPT. We recommend using an API model like [GPT-5](/api/docs/models/gpt-5) or [GPT-4o](/api/docs/models/gpt-4o) for most API integrations, but feel free to use this ChatGPT-4o model to test our latest improvements for chat use cases.`,status:`deprecated`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:3,speed:3,tagline:`GPT-4o model used in ChatGPT`,capabilities:{streaming:!0,vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[5,15]}]}],input:5,output:15},endpoints:[`chat_completions`,`responses`],snapshots:[`chatgpt-4o-latest`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/chatgpt-4o-latest`,open_weight:!1},{id:`chatgpt-image-latest`,name:`chatgpt-image-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`chatgpt`,description:`GPT Image Latest points to the image snapshot currently used in ChatGPT.`,status:`active`,model_type:`image`,performance:5,speed:3,tagline:`Image model used in ChatGPT.`,modalities:{input:[`text`,`image`],output:[`image`,`text`]},pricing:{input:5,output:10,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,10]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,2,32]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.009,.013,.013]},{label:`Medium`,values:[null,.034,.05,.05]},{label:`High`,values:[null,.133,.2,.2]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`],snapshots:[`chatgpt-image-latest`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/chatgpt-image-latest`,open_weight:!1},{id:`Code Interpreter`,name:`Code Interpreter`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,pricing:{tiers:[{label:`Built-in tools`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.03]}]}]},license:`proprietary`,page_url:`https://platform.openai.com/docs/models/Code Interpreter`,open_weight:!1},{id:`codex-mini-latest`,name:`codex-mini-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`codex`,description:`codex-mini-latest is a fine-tuned version of o4-mini specifically`,status:`deprecated`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Fast reasoning model optimized for the Codex CLI`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.5,output:6,cached_input:.375,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.5,.375,6]}]}]},endpoints:[`responses`],snapshots:[`codex-mini-latest`],model_type:`code`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/codex-mini-latest`,open_weight:!1,tools:[`function_calling`]},{id:`computer-use-preview-2025-03-11`,name:`computer-use-preview-2025-03-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`The computer-use-preview model is a specialized model for the computer use`,status:`active`,knowledge_cutoff:`2023-09`,context_window:8192,max_output_tokens:1024,reasoning_tokens:!0,alias:`computer-use-preview`,performance:2,reasoning:2,speed:2,tagline:`The computer-use-preview model is a specialized model for the computer use`,capabilities:{tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:3,output:12,batch_input:1.5,batch_output:6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,null,12]},{label:`Batch`,values:[1.5,null,6]}]}]},endpoints:[`responses`,`batch`],license:`proprietary`,release_date:`2025-03-11`,page_url:`https://platform.openai.com/docs/models/computer-use-preview`,open_weight:!1,tools:[`function_calling`],model_type:`chat`},{id:`computer-use-preview`,name:`computer-use-preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`The computer-use-preview model is a specialized model for the computer use`,status:`active`,knowledge_cutoff:`2023-09`,context_window:8192,max_output_tokens:1024,reasoning_tokens:!0,performance:2,reasoning:2,speed:2,tagline:`Specialized model for computer use tool`,capabilities:{tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:3,output:12,batch_input:1.5,batch_output:6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,null,12]},{label:`Batch`,values:[1.5,null,6]}]}]},endpoints:[`responses`,`batch`],snapshots:[`computer-use-preview-2025-03-11`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/computer-use-preview`,open_weight:!1,tools:[`function_calling`],release_date:`2025-03-11`,model_type:`chat`},{id:`dall-e-2`,name:`DALL·E 2`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`DALL·E is an AI system that creates realistic images and art from a natural language description. Older than DALL·E 3, DALL·E 2 offers more control in prompting and more requests at once.`,status:`deprecated`,model_type:`image`,performance:1,speed:2,tagline:`Our first image generation model`,successor:`dall-e-3`,modalities:{input:[`text`],output:[`image`]},pricing:{tiers:[{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`256x256`,`512x512`,`1024x1024`],rows:[{label:`Standard`,values:[null,.016,.018,.02]}]}]},endpoints:[`image_generation`,`image_edit`],snapshots:[`dall-e-2`],family:`dall-e`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/dall-e-2`,open_weight:!1},{id:`dall-e-3`,name:`DALL·E 3`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`DALL·E is an AI system that creates realistic images and art from a natural language description. DALL·E 3 currently supports the ability, given a prompt, to create a new image with a specific size.`,status:`deprecated`,model_type:`image`,performance:3,speed:2,tagline:`Previous generation image generation model`,modalities:{input:[`text`],output:[`image`]},pricing:{tiers:[{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1792`,`1792x1024`],rows:[{label:`Standard`,values:[null,.04,.08,.08]},{label:`HD`,values:[null,.08,.12,.12]}]}]},endpoints:[`image_generation`],snapshots:[`dall-e-3`],family:`dall-e`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/dall-e-3`,open_weight:!1},{id:`davinci-002`,name:`davinci-002`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`GPT base models can understand and generate natural language or code but are not trained with instruction following. These models are made to be replacements for our original GPT-3 base models and use the legacy Completions API. Most customers should use GPT-3.5 or GPT-4.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:16384,model_type:`chat`,performance:1,speed:3,tagline:`Replacement for the GPT-3 curie and davinci base models`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[6,12,null,12]},{label:`Batch`,values:[null,6,null,6]}]},{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[2,2]},{label:`Batch`,values:[1,1]}]}],input:2,output:2,batch_input:1,batch_output:1},endpoints:[`completions`],snapshots:[`davinci-002`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/davinci-002`,open_weight:!1},{id:`File Search Storage`,name:`File Search Storage`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,pricing:{tiers:[{label:`Built-in tools`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.1]}]}]},model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/File Search Storage`,open_weight:!1},{id:`gpt-3.5-0301`,name:`gpt-3.5-0301`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-3.5`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,performance:1,speed:2,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-3.5-0301`,open_weight:!1},{id:`gpt-3.5-turbo-0125`,name:`gpt-3.5-turbo-0125`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-3.5`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,performance:1,speed:2,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-3.5-turbo-0125`,open_weight:!1},{id:`gpt-3.5-turbo-0613`,name:`gpt-3.5-turbo-0613`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-3.5`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,performance:1,speed:2,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`batch`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-3.5-turbo-0613`,open_weight:!1},{id:`gpt-3.5-turbo-16k-0613`,name:`gpt-3.5-turbo-16k-0613`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-3.5`,description:`GPT-3.5 Turbo models can understand and generate natural language or code and have been optimized for chat using the Chat Completions API but work well for non-chat tasks as well. As of July 2024, use gpt-4o-mini in place of GPT-3.5 Turbo, as it is cheaper, more capable, multimodal, and just as fast. GPT-3.5 Turbo is still available for use in the API.`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,model_type:`chat`,performance:1,speed:2,tagline:`Legacy GPT model for cheaper chat and non-chat tasks`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[3,4]},{label:`Batch`,values:[1.5,2]}]}],input:3,output:4,batch_input:1.5,batch_output:2},endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-3.5-turbo-16k-0613`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-3-5-turbo-16k-0613`,open_weight:!1},{id:`gpt-3.5-turbo-instruct`,name:`gpt-3.5-turbo-instruct`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-3.5`,description:`Similar capabilities as GPT-3 era models. Compatible with legacy Completions endpoint and not Chat Completions.`,status:`deprecated`,knowledge_cutoff:`2021-08`,context_window:4096,max_output_tokens:4096,model_type:`chat`,performance:1,speed:2,tagline:`An older model only compatible with the legacy Completions endpoint`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[1.5,2]}]}],input:1.5,output:2},endpoints:[`chat_completions`,`responses`],snapshots:[`gpt-3.5-turbo-instruct`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-3-5-turbo-instruct`,open_weight:!1},{id:`gpt-3.5-turbo`,name:`GPT-3.5 Turbo`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-3.5`,description:`GPT-3.5 Turbo models can understand and generate natural language or code and have been optimized for chat using the Chat Completions API but work well for non-chat tasks as well. As of July 2024, use gpt-4o-mini in place of GPT-3.5 Turbo, as it is cheaper, more capable, multimodal, and just as fast. GPT-3.5 Turbo is still available for use in the API.`,status:`active`,knowledge_cutoff:`2021-08`,context_window:16385,max_output_tokens:4096,model_type:`chat`,performance:1,speed:2,tagline:`Legacy GPT model for cheaper chat and non-chat tasks`,successor:`gpt-4o-mini`,capabilities:{fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,3,null,6]},{label:`Batch`,values:[null,1.5,null,3]}]},{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[.5,1.5]},{label:`Batch`,values:[.25,.75]}]}],input:.5,output:1.5,batch_input:.25,batch_output:.75},endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`],snapshots:[`gpt-3.5-turbo-0125`,`gpt-3.5-turbo-1106`,`gpt-3.5-turbo-instruct`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-3-5-turbo`,open_weight:!1},{id:`gpt-4-0613`,name:`gpt-4-0613`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4`,status:`active`,knowledge_cutoff:`2023-11`,context_window:8192,max_output_tokens:8192,performance:2,speed:3,capabilities:{streaming:!0,fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4-0613`,open_weight:!1},{id:`gpt-4-1106-vision-preview`,name:`gpt-4-1106-vision-preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4`,status:`active`,knowledge_cutoff:`2023-11`,context_window:128e3,max_output_tokens:4096,performance:2,speed:3,capabilities:{streaming:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4-1106-vision-preview`,open_weight:!1},{id:`gpt-4-32k`,name:`gpt-4-32k`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4`,status:`active`,pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[60,120]},{label:`Batch`,values:[30,60]}]}],input:60,output:120,batch_input:30,batch_output:60},model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4-32k`,open_weight:!1},{id:`gpt-4-turbo-2024-04-09`,name:`gpt-4-turbo-2024-04-09`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4-turbo`,status:`active`,knowledge_cutoff:`2023-11`,context_window:128e3,max_output_tokens:4096,alias:`gpt-4-turbo`,performance:2,speed:3,capabilities:{streaming:!0,vision:!0,tool_call:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],description:`GPT-4 Turbo is the next generation of GPT-4, an older high-intelligence GPT model. It was designed to be a cheaper, better version of GPT-4. Today, we recommend using a newer model like GPT-4o.`,model_type:`chat`,tagline:`GPT-4 Turbo is the next generation of GPT-4, an older high-intelligence GPT model.`,pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[10,30]},{label:`Batch`,values:[5,15]}]}],input:10,output:30,batch_input:5,batch_output:15},license:`proprietary`,release_date:`2024-04-09`,page_url:`https://platform.openai.com/docs/models/gpt-4-turbo`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4-turbo-preview`,name:`GPT-4 Turbo Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4-turbo`,description:`This is a research preview of the GPT-4 Turbo model, an older high-intelligence GPT model.`,status:`deprecated`,model_type:`chat`,tagline:`An older fast GPT model`,successor:`gpt-4o`,snapshots:[`gpt-4-0125-preview`,`gpt-4-1106-vision-preview`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4-turbo-preview`,open_weight:!1},{id:`gpt-4-turbo`,name:`GPT-4 Turbo`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4-turbo`,status:`active`,knowledge_cutoff:`2023-11`,context_window:128e3,max_output_tokens:4096,performance:2,speed:3,capabilities:{streaming:!0,vision:!0,tool_call:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`gpt-4-turbo-2024-04-09`],description:`GPT-4 Turbo is the next generation of GPT-4, an older high-intelligence GPT model. It was designed to be a cheaper, better version of GPT-4. Today, we recommend using a newer model like GPT-4o.`,model_type:`chat`,tagline:`An older high-intelligence GPT model`,successor:`gpt-4o`,pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[10,30]},{label:`Batch`,values:[5,15]}]}],input:10,output:30,batch_input:5,batch_output:15},license:`proprietary`,release_date:`2024-04-09`,page_url:`https://platform.openai.com/docs/models/gpt-4-turbo`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4.1-2025-04-14`,name:`gpt-4.1-2025-04-14`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4.1`,description:`GPT-4.1 excels at instruction following and tool calling, with broad`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,alias:`gpt-4.1`,performance:4,speed:3,tagline:`GPT-4.1 excels at instruction following and tool calling, with broad`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3,.75,12]},{label:`Batch`,values:[null,1.5,null,6]}]}],input:2,output:8,cached_input:.5,batch_input:1,batch_output:4},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],license:`proprietary`,release_date:`2025-04-14`,page_url:`https://platform.openai.com/docs/models/gpt-4.1`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4.1-mini-2025-04-14`,name:`gpt-4.1-mini-2025-04-14`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4.1`,description:`GPT-4.1 mini excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,alias:`gpt-4.1-mini`,performance:3,speed:4,tagline:`GPT-4.1 mini excels at instruction following and tool calling.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.8,.2,3.2]},{label:`Batch`,values:[null,.4,null,1.6]}]}],input:.4,output:1.6,cached_input:.1,batch_input:.2,batch_output:.8},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],license:`proprietary`,release_date:`2025-04-14`,page_url:`https://platform.openai.com/docs/models/gpt-4.1-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4.1-mini`,name:`GPT-4.1 mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4.1`,description:`GPT-4.1 mini excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,performance:3,speed:4,tagline:`Smaller, faster version of GPT-4.1`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.4,output:1.6,cached_input:.1,batch_input:.2,batch_output:.8,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.8,.2,3.2]},{label:`Batch`,values:[null,.4,null,1.6]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4.1-mini-2025-04-14`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4.1-mini`,open_weight:!1,release_date:`2025-04-14`},{id:`gpt-4.1-nano-2025-04-14`,name:`gpt-4.1-nano-2025-04-14`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4.1`,description:`GPT-4.1 nano excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,alias:`gpt-4.1-nano`,performance:2,speed:5,tagline:`GPT-4.1 nano excels at instruction following and tool calling.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.5,.2,.05,.8]},{label:`Batch`,values:[null,.1,null,.4]}]}],input:.1,output:.4,cached_input:.025,batch_input:.05,batch_output:.2},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],license:`proprietary`,release_date:`2025-04-14`,page_url:`https://platform.openai.com/docs/models/gpt-4.1-nano`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4.1-nano`,name:`GPT-4.1 nano`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4.1`,description:`GPT-4.1 nano excels at instruction following and tool calling. It features a`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,performance:2,speed:5,tagline:`Fastest, most cost-efficient version of GPT-4.1`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.1,output:.4,cached_input:.025,batch_input:.05,batch_output:.2,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.5,.2,.05,.8]},{label:`Batch`,values:[null,.1,null,.4]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4.1-nano-2025-04-14`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4.1-nano`,open_weight:!1,release_date:`2025-04-14`},{id:`gpt-4.1`,name:`GPT-4.1`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4.1`,description:`GPT-4.1 excels at instruction following and tool calling, with broad`,status:`active`,knowledge_cutoff:`2024-05`,context_window:1047576,max_output_tokens:32768,model_type:`chat`,performance:4,speed:3,tagline:`Smartest non-reasoning model`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3,.75,12]},{label:`Batch`,values:[null,1.5,null,6]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4.1-2025-04-14`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4.1`,open_weight:!1,release_date:`2025-04-14`},{id:`gpt-4.5-preview-2025-02-27`,name:`gpt-4.5-preview-2025-02-27`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`Deprecated - a research preview of GPT-4.5. We recommend using gpt-4.1 or o3`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4.5-preview`,performance:4,speed:3,tagline:`Deprecated - a research preview of GPT-4.5.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:75,output:150,cached_input:37.5,batch_input:37.5,batch_output:75,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[75,37.5,150]},{label:`Batch`,values:[37.5,null,75]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],family:`gpt-4.5`,license:`proprietary`,release_date:`2025-02-27`,page_url:`https://platform.openai.com/docs/models/gpt-4.5-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4.5-preview`,name:`GPT-4.5 Preview (Deprecated)`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`Deprecated - a research preview of GPT-4.5. We recommend using gpt-4.1 or o3`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:3,tagline:`Deprecated large model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:75,output:150,cached_input:37.5,batch_input:37.5,batch_output:75,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[75,37.5,150]},{label:`Batch`,values:[37.5,null,75]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`gpt-4.5-preview-2025-02-27`],family:`gpt-4.5`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4.5-preview`,open_weight:!1,tools:[`function_calling`],release_date:`2025-02-27`},{id:`gpt-4`,name:`GPT-4`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4`,description:`GPT-4 is an older version of a high-intelligence GPT model, usable in Chat Completions.`,status:`active`,knowledge_cutoff:`2023-11`,context_window:8192,max_output_tokens:8192,model_type:`chat`,performance:2,speed:3,tagline:`An older high-intelligence GPT model`,successor:`gpt-4o`,capabilities:{streaming:!0,fine_tuning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:``,columns:[`Input`,`Output`],rows:[{label:`Standard`,values:[30,60]},{label:`Batch`,values:[15,30]}]}],input:30,output:60,batch_input:15,batch_output:30},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4-0613`,`gpt-4-0314`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4`,open_weight:!1},{id:`gpt-4o-2024-05-13`,name:`gpt-4o-2024-05-13`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:4096,model_type:`chat`,alias:`gpt-4o`,performance:3,speed:3,tagline:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,1.25,10]},{label:`Batch`,values:[1.25,null,5]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],license:`proprietary`,release_date:`2024-05-13`,page_url:`https://platform.openai.com/docs/models/gpt-4o`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-2024-08-06`,name:`gpt-4o-2024-08-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o`,performance:3,speed:3,tagline:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3.75,1.875,15]},{label:`Batch`,values:[null,1.875,null,7.5]}]}],input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],license:`proprietary`,release_date:`2024-08-06`,page_url:`https://platform.openai.com/docs/models/gpt-4o`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-2024-11-20`,name:`gpt-4o-2024-11-20`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o`,performance:3,speed:3,tagline:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3.75,1.875,15]},{label:`Batch`,values:[null,1.875,null,7.5]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],license:`proprietary`,release_date:`2024-11-20`,page_url:`https://platform.openai.com/docs/models/gpt-4o`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-audio-preview-2024-10-01`,name:`gpt-4o-audio-preview-2024-10-01`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-audio-preview`,performance:3,speed:3,tagline:`This is a preview release of the GPT-4o Audio models.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`],license:`proprietary`,release_date:`2024-10-01`,page_url:`https://platform.openai.com/docs/models/gpt-4o-audio-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-audio-preview-2024-12-17`,name:`gpt-4o-audio-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-audio-preview`,performance:3,speed:3,tagline:`This is a preview release of the GPT-4o Audio models.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`],license:`proprietary`,release_date:`2024-12-17`,page_url:`https://platform.openai.com/docs/models/gpt-4o-audio-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-audio-preview-2025-06-03`,name:`gpt-4o-audio-preview-2025-06-03`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-audio-preview`,performance:3,speed:3,tagline:`This is a preview release of the GPT-4o Audio models.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`],license:`proprietary`,release_date:`2025-06-03`,page_url:`https://platform.openai.com/docs/models/gpt-4o-audio-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-audio-preview`,name:`GPT-4o Audio`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Audio models. These models accept`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:3,speed:3,tagline:`GPT-4o models capable of audio inputs and outputs`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,null,80]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-4o-audio-preview-2025-06-03`,`gpt-4o-audio-preview-2024-12-17`,`gpt-4o-audio-preview-2024-10-01`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-audio-preview`,open_weight:!1,tools:[`function_calling`],release_date:`2025-06-03`},{id:`gpt-4o-mini-2024-07-18`,name:`gpt-4o-mini-2024-07-18`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o mini (“o” for “omni”) is a fast, affordable small model for focused tasks.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-mini`,performance:2,speed:4,tagline:`GPT-4o mini (“o” for “omni”) is a fast, affordable small model for focused tasks.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,.3,.15,1.2]},{label:`Batch`,values:[null,.15,null,.6]}]}],input:.15,output:.6,cached_input:.075,batch_input:.075,batch_output:.3},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],license:`proprietary`,release_date:`2024-07-18`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-mini-audio-preview-2024-12-17`,name:`gpt-4o-mini-audio-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the smaller GPT-4o Audio mini model. It's designed to input audio or create audio outputs via the REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-4o-mini-audio-preview`,performance:2,speed:4,tagline:`This is a preview release of the smaller GPT-4o Audio mini model.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},endpoints:[`chat_completions`],license:`proprietary`,release_date:`2024-12-17`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-audio-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-mini-audio-preview`,name:`GPT-4o mini Audio`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the smaller GPT-4o Audio mini model. It's designed to input audio or create audio outputs via the REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:2,speed:4,tagline:`Smaller model capable of audio inputs and outputs`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`],snapshots:[`gpt-4o-mini-audio-preview-2024-12-17`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-audio-preview`,open_weight:!1,release_date:`2024-12-17`},{id:`gpt-4o-mini-realtime-preview-2024-12-17`,name:`gpt-4o-mini-realtime-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o-mini Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,alias:`gpt-4o-mini-realtime-preview`,performance:2,speed:5,tagline:`This is a preview release of the GPT-4o-mini Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.3,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.3,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]}]},endpoints:[`realtime`],model_type:`audio`,license:`proprietary`,release_date:`2024-12-17`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-realtime-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-mini-realtime-preview`,name:`GPT-4o mini Realtime`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o-mini Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,performance:2,speed:5,tagline:`Smaller realtime model for text and audio inputs and outputs`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.3,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.3,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]}]},endpoints:[`realtime`],snapshots:[`gpt-4o-mini-realtime-preview-2024-12-17`],model_type:`audio`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-realtime-preview`,open_weight:!1,tools:[`function_calling`],release_date:`2024-12-17`},{id:`gpt-4o-mini-search-preview-2025-03-11`,name:`gpt-4o-mini-search-preview-2025-03-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o mini Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,alias:`gpt-4o-mini-search-preview`,performance:2,speed:4,tagline:`GPT-4o mini Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API.`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]}]},endpoints:[`chat_completions`],model_type:`chat`,license:`proprietary`,release_date:`2025-03-11`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-search-preview`,open_weight:!1},{id:`gpt-4o-mini-search-preview`,name:`GPT-4o mini Search Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o mini Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,performance:2,speed:4,tagline:`Fast, affordable small model for web search`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:.15,output:.6,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.15,null,.6]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-4o-mini-search-preview-2025-03-11`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-search-preview`,open_weight:!1,release_date:`2025-03-11`},{id:`gpt-4o-mini-transcribe-2025-03-20`,name:`gpt-4o-mini-transcribe-2025-03-20`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,alias:`gpt-4o-mini-transcribe`,performance:3,speed:4,modalities:{input:[`audio`,`text`],output:[`text`]},endpoints:[`transcription`,`realtime`],description:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,tagline:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[1.25,5,.003]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[3,null,.003]}]}],input:1.25,output:5},license:`proprietary`,release_date:`2025-03-20`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-transcribe`,open_weight:!1},{id:`gpt-4o-mini-transcribe-2025-12-15`,name:`gpt-4o-mini-transcribe-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,alias:`gpt-4o-mini-transcribe`,performance:3,speed:4,modalities:{input:[`audio`,`text`],output:[`text`]},endpoints:[`transcription`,`realtime`],description:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,tagline:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[1.25,5,.003]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[3,null,.003]}]}],input:1.25,output:5},license:`proprietary`,release_date:`2025-12-15`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-transcribe`,open_weight:!1},{id:`gpt-4o-mini-transcribe`,name:`GPT-4o mini Transcribe`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,performance:3,speed:4,modalities:{input:[`audio`,`text`],output:[`text`]},endpoints:[`transcription`,`realtime`],snapshots:[`gpt-4o-mini-transcribe-2025-03-20`,`gpt-4o-mini-transcribe-2025-12-15`],description:`GPT-4o mini Transcribe is a speech-to-text model that uses GPT-4o mini to transcribe audio.`,tagline:`Speech-to-text model powered by GPT-4o mini`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[1.25,5,.003]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[3,null,.003]}]}],input:1.25,output:5},license:`proprietary`,release_date:`2025-12-15`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-transcribe`,open_weight:!1},{id:`gpt-4o-mini-tts-2025-03-20`,name:`gpt-4o-mini-tts-2025-03-20`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,status:`active`,model_type:`tts`,alias:`gpt-4o-mini-tts`,performance:4,speed:4,modalities:{input:[`text`],output:[`audio`]},endpoints:[`speech_generation`],description:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model. Use it to convert text to natural sounding spoken text. The maximum number of input tokens is 2000.`,tagline:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[.6,null,.015]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[null,12,.015]}]}],input:.6},license:`proprietary`,release_date:`2025-03-20`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-tts`,open_weight:!1},{id:`gpt-4o-mini-tts-2025-12-15`,name:`gpt-4o-mini-tts-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,status:`active`,model_type:`tts`,alias:`gpt-4o-mini-tts`,performance:4,speed:4,modalities:{input:[`text`],output:[`audio`]},endpoints:[`speech_generation`],description:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model. Use it to convert text to natural sounding spoken text. The maximum number of input tokens is 2000.`,tagline:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model.`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[.6,null,.015]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[null,12,.015]}]}],input:.6},license:`proprietary`,release_date:`2025-12-15`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-tts`,open_weight:!1},{id:`gpt-4o-mini-tts`,name:`GPT-4o mini TTS`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,status:`active`,model_type:`tts`,performance:4,speed:4,modalities:{input:[`text`],output:[`audio`]},endpoints:[`speech_generation`],snapshots:[`gpt-4o-mini-tts-2025-03-20`,`gpt-4o-mini-tts-2025-12-15`],description:`GPT-4o mini TTS is a text-to-speech model built on GPT-4o mini, a fast and powerful language model. Use it to convert text to natural sounding spoken text. The maximum number of input tokens is 2000.`,tagline:`Text-to-speech model powered by GPT-4o mini`,pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[.6,null,.015]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[null,12,.015]}]}],input:.6},license:`proprietary`,release_date:`2025-12-15`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini-tts`,open_weight:!1},{id:`gpt-4o-mini`,name:`GPT-4o mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o mini (“o” for “omni”) is a fast, affordable small model for focused tasks.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:2,speed:4,tagline:`Fast, affordable small model for focused tasks`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.15,output:.6,cached_input:.075,batch_input:.075,batch_output:.3,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[3,.3,.15,1.2]},{label:`Batch`,values:[null,.15,null,.6]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`,`fine_tuning`],snapshots:[`gpt-4o-mini-2024-07-18`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-mini`,open_weight:!1,release_date:`2024-07-18`},{id:`gpt-4o-realtime-preview-2024-10-01`,name:`gpt-4o-realtime-preview-2024-10-01`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,alias:`gpt-4o-realtime-preview`,performance:2,speed:4,tagline:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`],model_type:`audio`,license:`proprietary`,release_date:`2024-10-01`,page_url:`https://platform.openai.com/docs/models/gpt-4o-realtime-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-realtime-preview-2024-12-17`,name:`gpt-4o-realtime-preview-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:16e3,max_output_tokens:4096,alias:`gpt-4o-realtime-preview`,performance:3,speed:4,tagline:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`],model_type:`audio`,license:`proprietary`,release_date:`2024-12-17`,page_url:`https://platform.openai.com/docs/models/gpt-4o-realtime-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-realtime-preview-2025-06-03`,name:`gpt-4o-realtime-preview-2025-06-03`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-4o-realtime-preview`,performance:3,speed:4,tagline:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`],model_type:`audio`,license:`proprietary`,release_date:`2025-06-03`,page_url:`https://platform.openai.com/docs/models/gpt-4o-realtime-preview`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-4o-realtime-preview`,name:`GPT-4o Realtime`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`This is a preview release of the GPT-4o Realtime model, capable of responding to audio and text inputs in realtime over WebRTC or a WebSocket interface.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,performance:3,speed:4,tagline:`Model capable of realtime text and audio inputs and outputs`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:5,output:20,cached_input:2.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,2.5,20]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[40,2.5,80]}]}]},endpoints:[`realtime`],snapshots:[`gpt-4o-realtime-preview-2025-06-03`,`gpt-4o-realtime-preview-2024-12-17`,`gpt-4o-realtime-preview-2024-10-01`],model_type:`audio`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-realtime-preview`,open_weight:!1,tools:[`function_calling`],release_date:`2025-06-03`},{id:`gpt-4o-search-preview-2025-03-11`,name:`gpt-4o-search-preview-2025-03-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,alias:`gpt-4o-search-preview`,performance:3,speed:3,tagline:`GPT-4o Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API.`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]}]},endpoints:[`chat_completions`],model_type:`chat`,license:`proprietary`,release_date:`2025-03-11`,page_url:`https://platform.openai.com/docs/models/gpt-4o-search-preview`,open_weight:!1},{id:`gpt-4o-search-preview`,name:`GPT-4o Search Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o Search Preview is a specialized model trained to understand and execute [web search](/api/docs/guides/tools-web-search?api-mode=chat) queries with the Chat Completions API. In addition to token fees, web search queries have a fee per tool call. Learn more in the [pricing](/api/docs/pricing) page.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,performance:3,speed:3,tagline:`GPT model for web search in Chat Completions`,capabilities:{streaming:!0,vision:!0,structured_output:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-4o-search-preview-2025-03-11`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-search-preview`,open_weight:!1,release_date:`2025-03-11`},{id:`gpt-4o-transcribe-diarize`,name:`GPT-4o Transcribe Diarize`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o Transcribe Diarize is an automatic speech recognition (ASR) model with built-in speaker diarization, meaning it associates audio segments with different speakers in a conversation. This model is only available in the Transcription API.`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,performance:4,speed:3,tagline:`Transcription model that identifies who's speaking when`,modalities:{input:[`text`,`audio`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[2.5,10,.006]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[6,null,.006]}]}],input:2.5,output:10},endpoints:[`transcription`],snapshots:[`gpt-4o-transcribe-diarize`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-transcribe-diarize`,open_weight:!1},{id:`gpt-4o-transcribe`,name:`GPT-4o Transcribe`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o Transcribe is a speech-to-text model that uses GPT-4o to transcribe audio.`,status:`active`,knowledge_cutoff:`2024-05`,context_window:16e3,max_output_tokens:2e3,model_type:`transcription`,performance:4,speed:3,tagline:`Speech-to-text model powered by GPT-4o`,modalities:{input:[`audio`,`text`],output:[`text`]},pricing:{tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[2.5,10,.006]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Output`,`Estimated cost`],rows:[{label:`Standard`,values:[6,null,.006]}]}],input:2.5,output:10},endpoints:[`transcription`,`realtime`],snapshots:[`gpt-4o-transcribe`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o-transcribe`,open_weight:!1},{id:`gpt-4o`,name:`GPT-4o`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-4o`,description:`GPT-4o (“o” for “omni”) is our versatile, high-intelligence flagship model.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:3,speed:3,tagline:`Fast, intelligent, flexible GPT model`,capabilities:{streaming:!0,vision:!0,tool_call:!0,fine_tuning:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:10,cached_input:1.25,batch_input:1.25,batch_output:5,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[25,3.75,1.875,15]},{label:`Batch`,values:[null,1.875,null,7.5]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`gpt-4o-2024-11-20`,`gpt-4o-2024-08-06`,`gpt-4o-2024-05-13`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-4o`,open_weight:!1,release_date:`2024-11-20`},{id:`gpt-5-2025-08-07`,name:`gpt-5-2025-08-07`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5 is our previous model for coding, reasoning, and agentic tasks across domains.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5`,performance:4,reasoning:4,speed:3,tagline:`GPT-5 is our previous model for coding, reasoning, and agentic tasks across domains.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},endpoints:[`chat_completions`,`responses`,`batch`],license:`proprietary`,release_date:`2025-08-07`,page_url:`https://platform.openai.com/docs/models/gpt-5`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5-chat-latest`,name:`GPT-5 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5 Chat points to the GPT-5 snapshot previously used in ChatGPT.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5 model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`responses`],snapshots:[`gpt-5-chat-latest`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5-chat-latest`,open_weight:!1},{id:`gpt-5-codex`,name:`GPT-5-Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5-Codex is a version of GPT-5 optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`A version of GPT-5 optimized for agentic coding in Codex`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5-codex`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5-codex`,open_weight:!1},{id:`gpt-5-latest`,name:`gpt-5-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,status:`active`,pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5-latest`,open_weight:!1},{id:`gpt-5-mini-2025-08-07`,name:`gpt-5-mini-2025-08-07`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5 mini is a faster, more cost-efficient version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5-mini`,performance:3,reasoning:3,speed:4,tagline:`GPT-5 mini is a faster, more cost-efficient version of GPT-5.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.25,output:2,cached_input:.025,batch_input:.125,batch_output:1,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.25,.025,2]},{label:`Batch`,values:[.125,.0125,1]},{label:`Flex`,values:[.125,.0125,1]},{label:`Priority`,values:[.45,.05,3.6]}]}]},endpoints:[`chat_completions`,`responses`,`batch`],license:`proprietary`,release_date:`2025-08-07`,page_url:`https://platform.openai.com/docs/models/gpt-5-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5-mini`,name:`GPT-5 mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5 mini is a faster, more cost-efficient version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,performance:3,reasoning:3,speed:4,tagline:`Near-frontier intelligence for cost sensitive, low latency, high volume workloads`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.25,output:2,cached_input:.025,batch_input:.125,batch_output:1,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.25,.025,2]},{label:`Batch`,values:[.125,.0125,1]},{label:`Flex`,values:[.125,.0125,1]},{label:`Priority`,values:[.45,.05,3.6]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-5-mini-2025-08-07`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5-mini`,open_weight:!1,release_date:`2025-08-07`},{id:`gpt-5-nano-2025-08-07`,name:`gpt-5-nano-2025-08-07`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5 Nano is our fastest, cheapest version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5-nano`,performance:2,reasoning:2,speed:5,tagline:`GPT-5 Nano is our fastest, cheapest version of GPT-5.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.05,output:.4,cached_input:.005,batch_input:.025,batch_output:.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.05,.005,.4]},{label:`Batch`,values:[.025,.0025,.2]},{label:`Flex`,values:[.025,.0025,.2]}]}]},endpoints:[`chat_completions`,`responses`,`batch`],license:`proprietary`,release_date:`2025-08-07`,page_url:`https://platform.openai.com/docs/models/gpt-5-nano`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5-nano`,name:`GPT-5 nano`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5 Nano is our fastest, cheapest version of GPT-5. It's great for`,status:`active`,knowledge_cutoff:`2024-05`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,performance:2,reasoning:2,speed:5,tagline:`Fastest, most cost-efficient version of GPT-5`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.05,output:.4,cached_input:.005,batch_input:.025,batch_output:.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.05,.005,.4]},{label:`Batch`,values:[.025,.0025,.2]},{label:`Flex`,values:[.025,.0025,.2]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-5-nano-2025-08-07`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5-nano`,open_weight:!1,release_date:`2025-08-07`},{id:`gpt-5-pro-2025-10-06`,name:`gpt-5-pro-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:272e3,reasoning_tokens:!0,alias:`gpt-5-pro`,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:120,batch_input:7.5,batch_output:60,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,null,120]},{label:`Batch`,values:[7.5,null,60]}]}]},endpoints:[`responses`,`batch`],model_type:`chat`,license:`proprietary`,release_date:`2025-10-06`,page_url:`https://platform.openai.com/docs/models/gpt-5-pro`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5-pro`,name:`gpt-5-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:272e3,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:120,batch_input:7.5,batch_output:60,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,null,120]},{label:`Batch`,values:[7.5,null,60]}]}]},endpoints:[`responses`,`batch`],snapshots:[`gpt-5-pro-2025-10-06`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5-pro`,open_weight:!1,tools:[`function_calling`],release_date:`2025-10-06`},{id:`gpt-5.1-2025-11-13`,name:`gpt-5.1-2025-11-13`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.1`,description:`GPT-5.1 is our flagship model for coding and agentic tasks with configurable reasoning and non-reasoning effort.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5.1`,performance:4,reasoning:4,speed:3,tagline:`GPT-5.1 is our flagship model for coding and agentic tasks with configurable reasoning and non-reasoning effort.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},endpoints:[`responses`,`chat_completions`,`batch`],license:`proprietary`,release_date:`2025-11-13`,page_url:`https://platform.openai.com/docs/models/gpt-5.1`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5.1-chat-latest`,name:`GPT-5.1 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.1`,description:`GPT-5.1 Chat points to the GPT-5.1 snapshot currently used in ChatGPT.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5.1 model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`responses`,`chat_completions`],snapshots:[`gpt-5.1-chat-latest`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.1-chat-latest`,open_weight:!1},{id:`gpt-5.1-codex-max`,name:`GPT-5.1-Codex-Max`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.1`,description:`GPT‑5.1-Codex-Max is purpose-built for agentic coding.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:4,tagline:`A version of GPT-5.1-codex optimized for long running tasks.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5.1-codex-max`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.1-codex-max`,open_weight:!1},{id:`gpt-5.1-codex-mini`,name:`GPT-5.1 Codex mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.1`,description:`GPT-5.1 Codex mini is a smaller, more cost-effective, less-capable version of GPT-5.1-Codex.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Smaller, more cost-effective, less-capable version of GPT-5.1-Codex`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.25,output:2,cached_input:.025,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.25,.025,2]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5.1-codex-mini`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.1-codex-mini`,open_weight:!1},{id:`gpt-5.1-codex`,name:`GPT-5.1 Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.1`,description:`GPT-5.1-Codex is a version of GPT-5 optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`A version of GPT-5.1 optimized for agentic coding in Codex.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},tools:[`function_calling`,`web_search`],endpoints:[`responses`],snapshots:[`gpt-5.1-codex`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.1-codex`,open_weight:!1},{id:`gpt-5.1-latest`,name:`gpt-5.1-latest`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.1`,status:`active`,pricing:{input:1.25,output:10,cached_input:.125,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]}]}]},model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.1-latest`,open_weight:!1},{id:`gpt-5.1`,name:`GPT-5.1`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.1`,description:`GPT-5.1 is our flagship model for coding and agentic tasks with configurable reasoning and non-reasoning effort.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,model_type:`chat`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`The best model for coding and agentic tasks with configurable reasoning effort`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},tools:[`function_calling`],endpoints:[`responses`,`chat_completions`,`batch`],snapshots:[`gpt-5.1-2025-11-13`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.1`,open_weight:!1,release_date:`2025-11-13`},{id:`gpt-5.2-2025-12-11`,name:`gpt-5.2-2025-12-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.2`,description:`GPT-5.2 is our previous frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5.2`,performance:5,reasoning:5,speed:3,tagline:`GPT-5.2 is our previous frontier model for complex professional work.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,batch_input:.875,batch_output:7,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]},{label:`Batch`,values:[.875,.0875,7]},{label:`Flex`,values:[.875,.0875,7]},{label:`Priority`,values:[3.5,.35,28]}]}]},endpoints:[`responses`,`chat_completions`,`batch`],license:`proprietary`,release_date:`2025-12-11`,page_url:`https://platform.openai.com/docs/models/gpt-5.2`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5.2-chat-latest`,name:`GPT-5.2 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.2`,description:`GPT-5.2 Chat points to the GPT-5.2 snapshot currently used in ChatGPT.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5.2 model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`responses`,`chat_completions`],snapshots:[`gpt-5.2-chat-latest`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.2-chat-latest`,open_weight:!1},{id:`gpt-5.2-codex`,name:`GPT-5.2-Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.2`,description:"GPT-5.2-Codex is an upgraded version of GPT-5.2 optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.\\nGPT-5.2-Codex supports `low`, `medium`, `high`, and `xhigh` reasoning effort settings.\\nIf you want to learn more about prompting GPT-5.2-Codex, refer to our [dedicated guide](/cookbook/examples/gpt-5/codex_prompting_guide).\\n",status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Our most intelligent coding model optimized for long-horizon, agentic coding tasks.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`hosted_shell`,`skills`],endpoints:[`responses`],snapshots:[`gpt-5.2-codex`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.2-codex`,open_weight:!1},{id:`gpt-5.2-pro-2025-12-11`,name:`gpt-5.2-pro-2025-12-11`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.2`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,reasoning_tokens:!0,alias:`gpt-5.2-pro`,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:21,output:168,batch_input:10.5,batch_output:84,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[21,null,168]},{label:`Batch`,values:[10.5,null,84]},{label:`Flex`,values:[10.5,null,84]}]}]},endpoints:[`responses`],model_type:`chat`,license:`proprietary`,release_date:`2025-12-11`,page_url:`https://platform.openai.com/docs/models/gpt-5.2-pro`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5.2-pro`,name:`gpt-5.2-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.2`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:21,output:168,batch_input:10.5,batch_output:84,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[21,null,168]},{label:`Batch`,values:[10.5,null,84]},{label:`Flex`,values:[10.5,null,84]}]}]},endpoints:[`responses`],snapshots:[`gpt-5.2-pro-2025-12-11`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.2-pro`,open_weight:!1,tools:[`function_calling`],release_date:`2025-12-11`},{id:`gpt-5.2`,name:`GPT-5.2`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.2`,description:`GPT-5.2 is our previous frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,model_type:`chat`,reasoning_tokens:!0,performance:5,reasoning:5,speed:3,tagline:`Previous frontier model for professional work with configurable reasoning effort`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,batch_input:.875,batch_output:7,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]},{label:`Batch`,values:[.875,.0875,7]},{label:`Flex`,values:[.875,.0875,7]},{label:`Priority`,values:[3.5,.35,28]}]}]},tools:[`function_calling`],endpoints:[`responses`,`chat_completions`,`batch`],snapshots:[`gpt-5.2-2025-12-11`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.2`,open_weight:!1,release_date:`2025-12-11`},{id:`gpt-5.3-chat-latest`,name:`GPT-5.3 Chat`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.3`,description:`GPT-5.3 Chat points to the GPT-5.3 Instant snapshot currently used in ChatGPT.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:128e3,max_output_tokens:16384,max_input_tokens:272e3,model_type:`chat`,performance:3,speed:3,tagline:`GPT-5.3 Instant model used in ChatGPT`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`file_search`,`image_generation`,`code_interpreter`,`mcp`],endpoints:[`responses`,`chat_completions`],snapshots:[`gpt-5.3-chat-latest`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.3-chat-latest`,open_weight:!1},{id:`gpt-5.3-codex`,name:`GPT-5.3-Codex`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.3`,description:"GPT-5.3-Codex is optimized for agentic coding tasks in [Codex](https://developers.openai.com/codex) or similar environments.\\nGPT-5.3-Codex supports `low`, `medium`, `high`, and `xhigh` reasoning effort settings.\\nIf you want to learn more about prompting GPT-5.3-Codex, refer to our [dedicated guide](/cookbook/examples/gpt-5/codex_prompting_guide).\\n",status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`The most capable agentic coding model to date.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.75,output:14,cached_input:.175,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.75,.175,14]}]}]},tools:[`function_calling`,`web_search`,`hosted_shell`,`skills`],endpoints:[`responses`],snapshots:[`gpt-5.3-codex`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.3-codex`,open_weight:!1},{id:`gpt-5.4-2026-03-05`,name:`gpt-5.4-2026-03-05`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,description:`GPT-5.4 is our frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5.4`,performance:5,reasoning:5,speed:3,tagline:`GPT-5.4 is our frontier model for complex professional work.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:15,cached_input:.25,batch_input:1.25,batch_output:7.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,.25,15]},{label:`Batch`,values:[1.25,.13,7.5]},{label:`Flex`,values:[1.25,.13,7.5]},{label:`Priority`,values:[5,.5,30]}]}]},endpoints:[`responses`,`chat_completions`,`batch`],license:`proprietary`,release_date:`2026-03-05`,page_url:`https://platform.openai.com/docs/models/gpt-5.4`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5.4-mini-2026-03-17`,name:`gpt-5.4-mini-2026-03-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,description:`GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5.4-mini`,performance:4,reasoning:4,speed:4,tagline:`GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.75,output:4.5,cached_input:.075,batch_input:.375,batch_output:2.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.75,.075,4.5]},{label:`Batch`,values:[.375,.0375,2.25]},{label:`Flex`,values:[.375,.0375,2.25]}]}]},endpoints:[`responses`,`chat_completions`,`batch`],license:`proprietary`,release_date:`2026-03-17`,page_url:`https://platform.openai.com/docs/models/gpt-5.4-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5.4-mini`,name:`GPT-5.4 mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,description:`GPT-5.4 mini brings the strengths of GPT-5.4 to a faster, more efficient`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,performance:4,reasoning:4,speed:4,tagline:`Our strongest mini model yet for coding, computer use, and subagents`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.75,output:4.5,cached_input:.075,batch_input:.375,batch_output:2.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.75,.075,4.5]},{label:`Batch`,values:[.375,.0375,2.25]},{label:`Flex`,values:[.375,.0375,2.25]}]}]},tools:[`function_calling`],endpoints:[`responses`,`chat_completions`,`batch`],pricing_notes:[`Regional processing (data residency) endpoints are charged a 10% uplift for GPT-5.4 mini.`],snapshots:[`gpt-5.4-mini-2026-03-17`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.4-mini`,open_weight:!1,release_date:`2026-03-17`},{id:`gpt-5.4-nano-2026-03-17`,name:`gpt-5.4-nano-2026-03-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,description:`GPT-5.4 nano is designed for tasks where speed and cost matter most like`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,alias:`gpt-5.4-nano`,performance:3,reasoning:3,speed:4,tagline:`GPT-5.4 nano is designed for tasks where speed and cost matter most like`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.2,output:1.25,cached_input:.02,batch_input:.1,batch_output:.625,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.2,.02,1.25]},{label:`Batch`,values:[.1,.01,.625]},{label:`Flex`,values:[.1,.01,.625]}]}]},endpoints:[`responses`,`chat_completions`,`batch`],license:`proprietary`,release_date:`2026-03-17`,page_url:`https://platform.openai.com/docs/models/gpt-5.4-nano`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5.4-nano`,name:`GPT-5.4 nano`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,description:`GPT-5.4 nano is designed for tasks where speed and cost matter most like`,status:`active`,knowledge_cutoff:`2025-08`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,performance:3,reasoning:3,speed:4,tagline:`Our cheapest GPT-5.4-class model for simple high-volume tasks`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:.2,output:1.25,cached_input:.02,batch_input:.1,batch_output:.625,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.2,.02,1.25]},{label:`Batch`,values:[.1,.01,.625]},{label:`Flex`,values:[.1,.01,.625]}]}]},tools:[`function_calling`],endpoints:[`responses`,`chat_completions`,`batch`],pricing_notes:[`Regional processing (data residency) endpoints are charged a 10% uplift for GPT-5.4 nano.`],snapshots:[`gpt-5.4-nano-2026-03-17`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.4-nano`,open_weight:!1,release_date:`2026-03-17`},{id:`gpt-5.4-pro-2026-03-05`,name:`gpt-5.4-pro-2026-03-05`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,reasoning_tokens:!0,alias:`gpt-5.4-pro`,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:30,output:180,batch_input:15,batch_output:90,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[30,null,180]},{label:`Batch`,values:[15,null,90]},{label:`Flex`,values:[15,null,90]}]}]},endpoints:[`responses`],model_type:`chat`,license:`proprietary`,release_date:`2026-03-05`,page_url:`https://platform.openai.com/docs/models/gpt-5.4-pro`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-5.4-pro`,name:`gpt-5.4-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,capabilities:{streaming:!0,vision:!0,tool_call:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:30,output:180,batch_input:15,batch_output:90,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[30,null,180]},{label:`Batch`,values:[15,null,90]},{label:`Flex`,values:[15,null,90]}]}]},endpoints:[`responses`],snapshots:[`gpt-5.4-pro-2026-03-05`],model_type:`chat`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.4-pro`,open_weight:!1,tools:[`function_calling`],release_date:`2026-03-05`},{id:`gpt-5.4`,name:`GPT-5.4`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5.4`,description:`GPT-5.4 is our frontier model for complex professional work.`,status:`active`,knowledge_cutoff:`2025-08`,context_window:105e4,max_output_tokens:128e3,model_type:`chat`,reasoning_tokens:!0,performance:5,reasoning:5,speed:3,tagline:`Best intelligence at scale for agentic, coding, and professional workflows`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2.5,output:15,cached_input:.25,batch_input:1.25,batch_output:7.5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,.25,15]},{label:`Batch`,values:[1.25,.13,7.5]},{label:`Flex`,values:[1.25,.13,7.5]},{label:`Priority`,values:[5,.5,30]}]}]},tools:[`function_calling`],endpoints:[`responses`,`chat_completions`,`batch`],pricing_notes:[`For models with a 1.05M context window (GPT-5.4 and GPT-5.4 pro), prompts with >272K input tokens are priced at 2x input and 1.5x output for the full session for standard, batch, and flex.`,`Regional processing (data residency) endpoints are charged a 10% uplift for GPT-5.4 and GPT-5.4 pro.`],snapshots:[`gpt-5.4-2026-03-05`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5.4`,open_weight:!1,release_date:`2026-03-05`},{id:`gpt-5`,name:`GPT-5`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-5`,description:`GPT-5 is our previous model for coding, reasoning, and agentic tasks across domains.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:4e5,max_output_tokens:128e3,max_input_tokens:272e3,model_type:`chat`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Previous intelligent reasoning model for coding and agentic tasks with configurable reasoning effort`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.25,output:10,cached_input:.125,batch_input:.625,batch_output:5,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.25,.125,10]},{label:`Batch`,values:[.625,.0625,5]},{label:`Flex`,values:[.625,.0625,5]},{label:`Priority`,values:[2.5,.25,20]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`gpt-5-2025-08-07`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-5`,open_weight:!1,release_date:`2025-08-07`},{id:`gpt-audio-1.5`,name:`gpt-audio-1.5`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-audio`,description:`The gpt-audio model is our first generally available audio model. It accepts`,status:`active`,knowledge_cutoff:`2024-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:3,tagline:`The best voice model for audio in, audio out with Chat Completions.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,null,64]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-audio-1.5`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-audio-1.5`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-audio-2025-08-28`,name:`gpt-audio-2025-08-28`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-audio`,description:`The gpt-audio model is our first generally available audio model. It accepts`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-audio`,performance:4,speed:3,tagline:`The gpt-audio model is our first generally available audio model.`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,null,64]}]}]},endpoints:[`chat_completions`],license:`proprietary`,release_date:`2025-08-28`,page_url:`https://platform.openai.com/docs/models/gpt-audio`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-audio-mini-2025-10-06`,name:`gpt-audio-mini-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-audio`,description:`A cost-efficient version of GPT Audio. It accepts audio inputs and outputs, and can be used in the Chat Completions REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-audio-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Audio.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,null,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},endpoints:[`chat_completions`],license:`proprietary`,release_date:`2025-10-06`,page_url:`https://platform.openai.com/docs/models/gpt-audio-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-audio-mini-2025-12-15`,name:`gpt-audio-mini-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-audio`,description:`A cost-efficient version of GPT Audio. It accepts audio inputs and outputs, and can be used in the Chat Completions REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,alias:`gpt-audio-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Audio.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,null,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},endpoints:[`chat_completions`],license:`proprietary`,release_date:`2025-12-15`,page_url:`https://platform.openai.com/docs/models/gpt-audio-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-audio-mini`,name:`gpt-audio-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-audio`,description:`A cost-efficient version of GPT Audio. It accepts audio inputs and outputs, and can be used in the Chat Completions REST API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Audio`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,null,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,null,20]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`],snapshots:[`gpt-audio-mini-2025-10-06`,`gpt-audio-mini-2025-12-15`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-audio-mini`,open_weight:!1,release_date:`2025-12-15`},{id:`gpt-audio`,name:`gpt-audio`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-audio`,description:`The gpt-audio model is our first generally available audio model. It accepts`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:16384,model_type:`chat`,performance:4,speed:3,tagline:`For audio inputs and outputs with Chat Completions API`,capabilities:{streaming:!0,tool_call:!0},modalities:{input:[`text`,`audio`],output:[`text`,`audio`]},pricing:{input:2.5,output:10,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,null,10]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,null,64]}]}]},endpoints:[`chat_completions`],snapshots:[`gpt-audio-2025-08-28`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-audio`,open_weight:!1,tools:[`function_calling`],release_date:`2025-08-28`},{id:`gpt-image-1-mini`,name:`gpt-image-1-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-image-1`,description:`A cost-efficient version of GPT Image 1. It is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.`,status:`active`,model_type:`image`,performance:4,speed:1,tagline:`A cost-efficient version of GPT Image 1`,modalities:{input:[`text`,`image`],output:[`image`]},pricing:{input:2,cached_input:.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.2,null]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2.5,.25,8]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1792`,`1792x1024`],rows:[{label:`Low`,values:[null,.005,.006,.006]},{label:`Medium`,values:[null,.011,.015,.015]},{label:`High`,values:[null,.036,.052,.052]}]}]},endpoints:[`batch`,`image_edit`,`image_generation`],snapshots:[`gpt-image-1-mini`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-image-1-mini`,open_weight:!1},{id:`gpt-image-1.5-2025-12-16`,name:`gpt-image-1.5-2025-12-16`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-image-1.5`,description:`GPT Image 1.5 is our latest image generation model, with better instruction following and adherence to prompts. Learn more in our [GPT Image 1.5 usage guide](/api/docs/guides/image-generation).`,status:`active`,model_type:`image`,alias:`gpt-image-1.5`,performance:5,speed:3,tagline:`GPT Image 1.5 is our latest image generation model, with better instruction following and adherence to prompts.`,modalities:{input:[`text`,`image`],output:[`image`,`text`]},pricing:{input:5,output:10,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,10]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,2,32]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.009,.013,.013]},{label:`Medium`,values:[null,.034,.05,.05]},{label:`High`,values:[null,.133,.2,.2]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`],license:`proprietary`,release_date:`2025-12-16`,page_url:`https://platform.openai.com/docs/models/gpt-image-1.5`,open_weight:!1},{id:`gpt-image-1.5`,name:`GPT Image 1.5`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-image-1.5`,description:`GPT Image 1.5 is our latest image generation model, with better instruction following and adherence to prompts. Learn more in our [GPT Image 1.5 usage guide](/api/docs/guides/image-generation).`,status:`active`,model_type:`image`,performance:5,speed:3,tagline:`State-of-the-art image generation model.`,modalities:{input:[`text`,`image`],output:[`image`,`text`]},pricing:{input:5,output:10,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,10]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[8,2,32]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.009,.013,.013]},{label:`Medium`,values:[null,.034,.05,.05]},{label:`High`,values:[null,.133,.2,.2]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`],snapshots:[`gpt-image-1.5-2025-12-16`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-image-1.5`,open_weight:!1,release_date:`2025-12-16`},{id:`gpt-image-1`,name:`GPT Image 1`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-image-1`,description:`GPT Image 1 is a natively multimodal language model that accepts both text and image inputs, and produces image outputs.`,status:`active`,model_type:`image`,performance:4,speed:1,tagline:`Our previous image generation model`,modalities:{input:[`text`,`image`],output:[`image`]},pricing:{input:5,cached_input:1.25,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,1.25,null]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,2.5,40]}]},{label:`Image generation`,unit:`Per image`,columns:[`Quality`,`1024x1024`,`1024x1536`,`1536x1024`],rows:[{label:`Low`,values:[null,.011,.016,.016]},{label:`Medium`,values:[null,.042,.063,.063]},{label:`High`,values:[null,.167,.25,.25]}]}]},endpoints:[`batch`,`image_generation`,`image_edit`,`responses`],snapshots:[`gpt-image-1`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-image-1`,open_weight:!1},{id:`gpt-oss-120b`,name:`gpt-oss-120b`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-oss`,description:`\\`,status:`active`,knowledge_cutoff:`2024-05`,context_window:131072,max_output_tokens:131072,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Most powerful open-weight model, fits into an H100 GPU`,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},tools:[`function_calling`,`code_interpreter`,`mcp`,`web_search`],endpoints:[`responses`,`batch`],snapshots:[`gpt-oss-120b`],license:`apache-2.0`,parameters:120,page_url:`https://platform.openai.com/docs/models/gpt-oss-120b`,open_weight:!0},{id:`gpt-oss-20b`,name:`gpt-oss-20b`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-oss`,description:`\\`,status:`active`,knowledge_cutoff:`2024-05`,context_window:131072,max_output_tokens:131072,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Medium-sized open-weight model for low latency`,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},tools:[`function_calling`,`code_interpreter`,`mcp`,`web_search`],endpoints:[`responses`,`batch`],snapshots:[`gpt-oss-20b`],license:`apache-2.0`,parameters:20,page_url:`https://platform.openai.com/docs/models/gpt-oss-20b`,open_weight:!0},{id:`gpt-realtime-1.5`,name:`gpt-realtime-1.5`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-realtime`,description:`GPT-Reatime-1.5 is our flagship audio model for voice agents & customer support.`,status:`active`,knowledge_cutoff:`2024-09`,context_window:32e3,max_output_tokens:4096,performance:5,speed:4,tagline:`The best voice model for audio in, audio out.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`,`image`],output:[`text`,`audio`]},pricing:{input:4,output:16,cached_input:.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[4,.4,16]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,.4,64]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.5,null]}]}]},endpoints:[`realtime`],snapshots:[`gpt-realtime-1.5`],model_type:`audio`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-realtime-1.5`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-realtime-2025-08-28`,name:`gpt-realtime-2025-08-28`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-realtime`,description:`This is our first general-availability realtime model, capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-realtime`,performance:5,speed:4,tagline:`This is our first general-availability realtime model, capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`,`image`],output:[`text`,`audio`]},pricing:{input:4,output:16,cached_input:.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[4,.4,16]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,.4,64]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.5,null]}]}]},endpoints:[`realtime`],model_type:`audio`,license:`proprietary`,release_date:`2025-08-28`,page_url:`https://platform.openai.com/docs/models/gpt-realtime`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-realtime-mini-2025-10-06`,name:`gpt-realtime-mini-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-realtime`,description:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-realtime-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`image`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.06,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.06,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.8,.08,null]}]}]},endpoints:[`realtime`],model_type:`audio`,license:`proprietary`,release_date:`2025-10-06`,page_url:`https://platform.openai.com/docs/models/gpt-realtime-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-realtime-mini-2025-12-15`,name:`gpt-realtime-mini-2025-12-15`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-realtime`,description:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,alias:`gpt-realtime-mini`,performance:4,speed:5,tagline:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,capabilities:{tool_call:!0},modalities:{input:[`text`,`image`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.06,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.06,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.8,.08,null]}]}]},endpoints:[`realtime`],model_type:`audio`,license:`proprietary`,release_date:`2025-12-15`,page_url:`https://platform.openai.com/docs/models/gpt-realtime-mini`,open_weight:!1,tools:[`function_calling`]},{id:`gpt-realtime-mini`,name:`gpt-realtime-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-realtime`,description:`A cost-efficient version of GPT Realtime - capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,performance:4,speed:5,tagline:`A cost-efficient version of GPT Realtime`,capabilities:{tool_call:!0},modalities:{input:[`text`,`image`,`audio`],output:[`text`,`audio`]},pricing:{input:.6,output:2.4,cached_input:.06,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.6,.06,2.4]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,.3,20]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[.8,.08,null]}]}]},endpoints:[`realtime`],snapshots:[`gpt-realtime-mini-2025-10-06`,`gpt-realtime-mini-2025-12-15`],model_type:`audio`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-realtime-mini`,open_weight:!1,tools:[`function_calling`],release_date:`2025-12-15`},{id:`gpt-realtime`,name:`gpt-realtime`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`gpt-realtime`,description:`This is our first general-availability realtime model, capable of responding to audio and text inputs in realtime over WebRTC, WebSocket, or SIP connections.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:32e3,max_output_tokens:4096,performance:5,speed:4,tagline:`Model capable of realtime text and audio inputs and outputs`,capabilities:{tool_call:!0},modalities:{input:[`text`,`audio`,`image`],output:[`text`,`audio`]},pricing:{input:4,output:16,cached_input:.4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[4,.4,16]}]},{label:`Audio tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[32,.4,64]}]},{label:`Image tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[5,.5,null]}]}]},endpoints:[`realtime`],snapshots:[`gpt-realtime-2025-08-28`],model_type:`audio`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/gpt-realtime`,open_weight:!1,tools:[`function_calling`],release_date:`2025-08-28`},{id:`o1-2024-12-17`,name:`o1-2024-12-17`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1`,performance:4,reasoning:4,speed:1,tagline:`The o1 series of models are trained with reinforcement learning to perform complex reasoning.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:60,cached_input:7.5,batch_input:7.5,batch_output:30,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,7.5,60]},{label:`Batch`,values:[7.5,null,30]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],license:`proprietary`,release_date:`2024-12-17`,page_url:`https://platform.openai.com/docs/models/o1`,open_weight:!1,tools:[`function_calling`]},{id:`o1-mini-2024-09-12`,name:`o1-mini-2024-09-12`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,description:`The o1 reasoning model is designed to solve hard problems across domains. o1-mini is a faster and more affordable reasoning model, but we recommend using the newer o3-mini model that features higher intelligence at the same latency and price as o1-mini.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:65536,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1-mini`,performance:3,reasoning:3,speed:2,tagline:`The o1 reasoning model is designed to solve hard problems across domains.`,capabilities:{streaming:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},endpoints:[`chat_completions`,`assistants`],license:`proprietary`,release_date:`2024-09-12`,page_url:`https://platform.openai.com/docs/models/o1-mini`,open_weight:!1},{id:`o1-mini`,name:`o1-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,description:`The o1 reasoning model is designed to solve hard problems across domains. o1-mini is a faster and more affordable reasoning model, but we recommend using the newer o3-mini model that features higher intelligence at the same latency and price as o1-mini.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:65536,model_type:`reasoning`,reasoning_tokens:!0,performance:3,reasoning:3,speed:2,tagline:`A small model alternative to o1`,successor:`o3-mini`,capabilities:{streaming:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},tools:[`file_search`,`code_interpreter`,`mcp`],endpoints:[`chat_completions`,`assistants`],snapshots:[`o1-mini-2024-09-12`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o1-mini`,open_weight:!1,release_date:`2024-09-12`},{id:`o1-preview-2024-09-12`,name:`o1-preview-2024-09-12`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,status:`active`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:32768,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1-preview`,performance:3,reasoning:3,speed:1,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`assistants`],description:`Research preview of the o1 series of models, trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,tagline:`Research preview of the o1 series of models, trained with reinforcement learning to perform complex reasoning.`,license:`proprietary`,release_date:`2024-09-12`,page_url:`https://platform.openai.com/docs/models/o1-preview`,open_weight:!1,tools:[`function_calling`]},{id:`o1-preview`,name:`o1 Preview`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,status:`deprecated`,knowledge_cutoff:`2023-09`,context_window:128e3,max_output_tokens:32768,model_type:`reasoning`,reasoning_tokens:!0,performance:3,reasoning:3,speed:1,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},endpoints:[`chat_completions`,`assistants`],snapshots:[`o1-preview-2024-09-12`],description:`Research preview of the o1 series of models, trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,tagline:`Preview of our first o-series reasoning model`,successor:`o1`,license:`proprietary`,release_date:`2024-09-12`,page_url:`https://platform.openai.com/docs/models/o1-preview`,open_weight:!1,tools:[`function_calling`]},{id:`o1-pro-2025-03-19`,name:`o1-pro-2025-03-19`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o1-pro`,performance:4,reasoning:4,speed:1,tagline:`The o1 series of models are trained with reinforcement learning to think`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:150,output:600,batch_input:75,batch_output:300,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[150,null,600]},{label:`Batch`,values:[75,null,300]}]}]},endpoints:[`responses`,`batch`],license:`proprietary`,release_date:`2025-03-19`,page_url:`https://platform.openai.com/docs/models/o1-pro`,open_weight:!1,tools:[`function_calling`]},{id:`o1-pro`,name:`o1-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:1,tagline:`Version of o1 with more compute for better responses`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:150,output:600,batch_input:75,batch_output:300,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[150,null,600]},{label:`Batch`,values:[75,null,300]}]}]},tools:[`function_calling`],endpoints:[`responses`,`batch`],snapshots:[`o1-pro-2025-03-19`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o1-pro`,open_weight:!1,release_date:`2025-03-19`},{id:`o1`,name:`o1`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o1`,description:`The o1 series of models are trained with reinforcement learning to perform complex reasoning. o1 models think before they answer, producing a long internal chain of thought before responding to the user.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:1,tagline:`Previous full o-series reasoning model`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:15,output:60,cached_input:7.5,batch_input:7.5,batch_output:30,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[15,7.5,60]},{label:`Batch`,values:[7.5,null,30]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`o1-2024-12-17`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o1`,open_weight:!1,release_date:`2024-12-17`},{id:`o3-2025-04-16`,name:`o3-2025-04-16`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`o3 is a well-rounded and powerful model across domains. It sets a new`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o3`,performance:5,reasoning:5,speed:1,tagline:`o3 is a well-rounded and powerful model across domains.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]},{label:`Flex`,values:[1,.25,4]}]}]},endpoints:[`chat_completions`,`responses`,`batch`],license:`proprietary`,release_date:`2025-04-16`,page_url:`https://platform.openai.com/docs/models/o3`,open_weight:!1,tools:[`function_calling`]},{id:`o3-deep-research-2025-06-26`,name:`o3-deep-research-2025-06-26`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`o3-deep-research is our most advanced model for deep research,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`chat`,reasoning_tokens:!0,alias:`o3-deep-research`,performance:5,reasoning:5,speed:1,tagline:`o3-deep-research is our most advanced model for deep research,`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:10,output:40,cached_input:2.5,batch_input:5,batch_output:20,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,2.5,40]},{label:`Batch`,values:[5,null,20]}]}]},endpoints:[`responses`,`batch`],license:`proprietary`,release_date:`2025-06-26`,page_url:`https://platform.openai.com/docs/models/o3-deep-research`,open_weight:!1},{id:`o3-deep-research`,name:`o3-deep-research`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`o3-deep-research is our most advanced model for deep research,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`chat`,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,tagline:`Our most powerful deep research model`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:10,output:40,cached_input:2.5,batch_input:5,batch_output:20,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[10,2.5,40]},{label:`Batch`,values:[5,null,20]}]}]},tools:[`web_search`,`code_interpreter`,`mcp`],endpoints:[`responses`,`batch`],snapshots:[`o3-deep-research-2025-06-26`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o3-deep-research`,open_weight:!1,release_date:`2025-06-26`},{id:`o3-mini-2025-01-31`,name:`o3-mini-2025-01-31`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`o3-mini is our newest small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini. o3-mini supports key developer features, like Structured Outputs, function calling, and Batch API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o3-mini`,performance:4,reasoning:4,speed:3,tagline:`o3-mini is our newest small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini.`,capabilities:{streaming:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],license:`proprietary`,release_date:`2025-01-31`,page_url:`https://platform.openai.com/docs/models/o3-mini`,open_weight:!1,tools:[`function_calling`]},{id:`o3-mini`,name:`o3-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`o3-mini is our newest small reasoning model, providing high intelligence at the same cost and latency targets of o1-mini. o3-mini supports key developer features, like Structured Outputs, function calling, and Batch API.`,status:`active`,knowledge_cutoff:`2023-09`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`A small model alternative to o3`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.55,batch_input:.55,batch_output:2.2,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[1.1,.55,4.4]},{label:`Batch`,values:[.55,null,2.2]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`assistants`,`batch`],snapshots:[`o3-mini-2025-01-31`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o3-mini`,open_weight:!1,release_date:`2025-01-31`},{id:`o3-pro-2025-06-10`,name:`o3-pro-2025-06-10`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`The o-series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o3-pro`,performance:5,reasoning:5,speed:1,tagline:`The o-series of models are trained with reinforcement learning to think`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:20,output:80,batch_input:10,batch_output:40,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[20,null,80]},{label:`Batch`,values:[10,null,40]}]}]},endpoints:[`responses`,`batch`],license:`proprietary`,release_date:`2025-06-10`,page_url:`https://platform.openai.com/docs/models/o3-pro`,open_weight:!1,tools:[`function_calling`]},{id:`o3-pro`,name:`o3-pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`The o-series of models are trained with reinforcement learning to think`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,tagline:`Version of o3 with more compute for better responses`,capabilities:{vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:20,output:80,batch_input:10,batch_output:40,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[20,null,80]},{label:`Batch`,values:[10,null,40]}]}]},tools:[`function_calling`],endpoints:[`responses`,`batch`],snapshots:[`o3-pro-2025-06-10`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o3-pro`,open_weight:!1,release_date:`2025-06-10`},{id:`o3`,name:`o3`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o3`,description:`o3 is a well-rounded and powerful model across domains. It sets a new`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:5,reasoning:5,speed:1,tagline:`Reasoning model for complex tasks, succeeded by GPT-5`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]},{label:`Flex`,values:[1,.25,4]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`batch`],snapshots:[`o3-2025-04-16`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o3`,open_weight:!1,release_date:`2025-04-16`},{id:`o4-mini-2025-04-16`,name:`o4-mini-2025-04-16`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o4`,description:`o4-mini is our latest small o-series model. It's optimized for fast,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,alias:`o4-mini`,performance:4,reasoning:4,speed:3,tagline:`o4-mini is our latest small o-series model.`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[null,4,1,16]},{label:`Batch`,values:[null,2,null,8]}]}],input:1.1,output:4.4,cached_input:.275,batch_input:.55,batch_output:2.2},endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`],license:`proprietary`,release_date:`2025-04-16`,page_url:`https://platform.openai.com/docs/models/o4-mini`,open_weight:!1,tools:[`function_calling`]},{id:`o4-mini-deep-research-2025-06-26`,name:`o4-mini-deep-research-2025-06-26`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o4`,description:`o4-mini-deep-research is our faster, more affordable deep`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`chat`,reasoning_tokens:!0,alias:`o4-mini-deep-research`,performance:4,reasoning:4,speed:3,tagline:`o4-mini-deep-research is our faster, more affordable deep`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]}]}]},endpoints:[`responses`,`batch`],license:`proprietary`,release_date:`2025-06-26`,page_url:`https://platform.openai.com/docs/models/o4-mini-deep-research`,open_weight:!1},{id:`o4-mini-deep-research`,name:`o4-mini-deep-research`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o4`,description:`o4-mini-deep-research is our faster, more affordable deep`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`chat`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Faster, more affordable deep research model`,capabilities:{streaming:!0,vision:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:2,output:8,cached_input:.5,batch_input:1,batch_output:4,tiers:[{label:`Text tokens`,unit:`Per 1M tokens`,columns:[`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[2,.5,8]},{label:`Batch`,values:[1,null,4]}]}]},tools:[`web_search`,`code_interpreter`,`mcp`],endpoints:[`responses`,`batch`],snapshots:[`o4-mini-deep-research-2025-06-26`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o4-mini-deep-research`,open_weight:!1,release_date:`2025-06-26`},{id:`o4-mini`,name:`o4-mini`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`o4`,description:`o4-mini is our latest small o-series model. It's optimized for fast,`,status:`active`,knowledge_cutoff:`2024-05`,context_window:2e5,max_output_tokens:1e5,model_type:`reasoning`,reasoning_tokens:!0,performance:4,reasoning:4,speed:3,tagline:`Fast, cost-efficient reasoning model, succeeded by GPT-5 mini`,capabilities:{streaming:!0,vision:!0,tool_call:!0,structured_output:!0,fine_tuning:!0,reasoning:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{input:1.1,output:4.4,cached_input:.275,batch_input:.55,batch_output:2.2,tiers:[{label:`Fine-tuning`,unit:``,columns:[`Training`,`Input`,`Cached input`,`Output`],rows:[{label:`Standard`,values:[null,4,1,16]},{label:`Batch`,values:[null,2,null,8]}]}]},tools:[`function_calling`],endpoints:[`chat_completions`,`responses`,`batch`,`fine_tuning`],snapshots:[`o4-mini-2025-04-16`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/o4-mini`,open_weight:!1,release_date:`2025-04-16`},{id:`omni-moderation-2024-09-26`,name:`omni-moderation-2024-09-26`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,model_type:`moderation`,alias:`omni-moderation`,performance:3,speed:3,capabilities:{vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`batch`,`moderation`],family:`omni-moderation`,license:`proprietary`,release_date:`2024-09-26`,page_url:`https://platform.openai.com/docs/models/omni-moderation`,open_weight:!1},{id:`omni-moderation-latest`,name:`omni-moderation`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`Moderation models are free models designed to detect harmful content.`,status:`active`,model_type:`moderation`,performance:3,speed:3,tagline:`Identify potentially harmful content in text and images`,capabilities:{vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},pricing:{tiers:[{label:`Moderation`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[null]}]}]},endpoints:[`batch`,`moderation`],snapshots:[`omni-moderation-2024-09-26`],family:`omni-moderation`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/omni-moderation-latest`,open_weight:!1},{id:`omni-moderation`,name:`omni-moderation`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,model_type:`moderation`,performance:3,speed:3,capabilities:{vision:!0},modalities:{input:[`text`,`image`],output:[`text`]},endpoints:[`batch`,`moderation`],snapshots:[`omni-moderation-2024-09-26`],family:`omni-moderation`,license:`proprietary`,release_date:`2024-09-26`,page_url:`https://platform.openai.com/docs/models/omni-moderation`,open_weight:!1},{id:`sora-2-2025-10-06`,name:`sora-2-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,model_type:`video`,alias:`sora-2`,performance:4,speed:2,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],description:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,tagline:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.1,null,null]}]}]},family:`sora-2`,license:`proprietary`,release_date:`2025-10-06`,page_url:`https://platform.openai.com/docs/models/sora-2`,open_weight:!1},{id:`sora-2-2025-12-08`,name:`sora-2-2025-12-08`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,model_type:`video`,alias:`sora-2`,performance:4,speed:2,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],description:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,tagline:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.1,null,null]}]}]},family:`sora-2`,license:`proprietary`,release_date:`2025-12-08`,page_url:`https://platform.openai.com/docs/models/sora-2`,open_weight:!1},{id:`sora-2-pro-2025-10-06`,name:`sora-2-pro-2025-10-06`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,model_type:`video`,alias:`sora-2-pro`,performance:5,speed:1,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],description:`Sora 2 Pro is our state-of-the-art, most advanced media generation model, generating videos with synced audio.`,tagline:`Sora 2 Pro is our state-of-the-art, most advanced media generation model, generating videos with synced audio.`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.3,.5,.7]}]}]},family:`sora-2`,license:`proprietary`,release_date:`2025-10-06`,page_url:`https://platform.openai.com/docs/models/sora-2-pro`,open_weight:!1},{id:`sora-2-pro`,name:`Sora 2 Pro`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,model_type:`video`,performance:5,speed:1,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],snapshots:[`sora-2-pro-2025-10-06`,`sora-2-pro`],description:`Sora 2 Pro is our state-of-the-art, most advanced media generation model, generating videos with synced audio.`,tagline:`Most advanced synced-audio video generation`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.3,.5,.7]}]}]},family:`sora-2`,license:`proprietary`,release_date:`2025-10-06`,page_url:`https://platform.openai.com/docs/models/sora-2-pro`,open_weight:!1},{id:`sora-2`,name:`Sora 2`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,model_type:`video`,performance:4,speed:2,modalities:{input:[`text`,`image`],output:[`video`,`audio`]},endpoints:[`videos`],snapshots:[`sora-2-2025-10-06`,`sora-2-2025-12-08`,`sora-2`],description:`Sora 2 is our new powerful media generation model, generating videos with synced audio.`,tagline:`Flagship video generation with synced audio`,pricing:{tiers:[{label:`Video generation`,unit:`Per second`,columns:[`Portrait: 720x1280 Landscape: 1280x720`,`Portrait: 1024x1792 Landscape: 1792x1024`,`Portrait: 1080x1920 Landscape: 1920x1080`],rows:[{label:`Standard`,values:[.1,null,null]}]}]},family:`sora-2`,license:`proprietary`,release_date:`2025-12-08`,page_url:`https://platform.openai.com/docs/models/sora-2`,open_weight:!1},{id:`text-embedding-3-large`,name:`text-embedding-3-large`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`text-embedding-3-large is our most capable embedding model for both english and non-english tasks.`,status:`active`,model_type:`embed`,performance:3,speed:2,tagline:`Most capable embedding model`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Embeddings`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.13]},{label:`Batch`,values:[.065]}]}]},endpoints:[`embeddings`,`batch`],snapshots:[`text-embedding-3-large`],family:`text-embedding`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/text-embedding-3-large`,open_weight:!1},{id:`text-embedding-3-small`,name:`text-embedding-3-small`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`text-embedding-3-small is our improved, more performant version of our ada embedding model.`,status:`active`,model_type:`embed`,performance:2,speed:3,tagline:`Small embedding model`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Embeddings`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.02]},{label:`Batch`,values:[.01]}]}]},endpoints:[`embeddings`,`batch`],snapshots:[`text-embedding-3-small`],family:`text-embedding`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/text-embedding-3-small`,open_weight:!1},{id:`text-embedding-ada-002`,name:`text-embedding-ada-002`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`text-embedding-ada-002 is our improved, more performant version of our ada embedding model.`,status:`active`,model_type:`embed`,performance:1,speed:2,tagline:`Older embedding model`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Embeddings`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[.1]},{label:`Batch`,values:[.05]}]}]},endpoints:[`embeddings`,`batch`],snapshots:[`text-embedding-ada-002`],family:`text-embedding`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/text-embedding-ada-002`,open_weight:!1},{id:`text-moderation-007`,name:`text-moderation-007`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,status:`active`,knowledge_cutoff:`2021-08`,max_output_tokens:32768,model_type:`moderation`,performance:2,speed:3,modalities:{input:[`text`],output:[`text`]},endpoints:[`moderation`],family:`text-moderation`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/text-moderation-007`,open_weight:!1},{id:`text-moderation-latest`,name:`text-moderation`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`Moderation models are free models designed to detect harmful content. This is our text only moderation model; we expect omni-moderation-* models to be the best default moving forward.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:32768,model_type:`moderation`,performance:2,speed:3,tagline:`Previous generation text-only moderation model`,successor:`omni-moderation`,modalities:{input:[`text`],output:[`text`]},pricing:{tiers:[{label:`Moderation`,unit:``,columns:[`Cost`],rows:[{label:`Standard`,values:[null]}]}]},endpoints:[`moderation`],snapshots:[`text-moderation-007`],family:`text-moderation`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/text-moderation-latest`,open_weight:!1},{id:`text-moderation-stable`,name:`text-moderation-stable`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`Moderation models are free models designed to detect harmful content. This is our text only moderation model; we expect omni-moderation-* models to be the best default moving forward.`,status:`deprecated`,knowledge_cutoff:`2021-08`,max_output_tokens:32768,model_type:`moderation`,performance:2,speed:3,tagline:`Previous generation text-only moderation model`,successor:`omni-moderation`,modalities:{input:[`text`],output:[`text`]},endpoints:[`moderation`],snapshots:[`text-moderation-007`],family:`text-moderation`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/text-moderation-stable`,open_weight:!1},{id:`tts-1-hd`,name:`TTS-1 HD`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`TTS is a model that converts text to natural sounding spoken text. The tts-1-hd model is optimized for high quality text-to-speech use cases. Use it with the Speech endpoint in the Audio API.`,status:`active`,model_type:`tts`,performance:3,speed:3,tagline:`Text-to-speech model optimized for quality`,modalities:{input:[`text`],output:[`audio`]},pricing:{tiers:[{label:`Pricing`,unit:`Per 1M tokens`,columns:[`Cost`],rows:[{label:`Standard`,values:[30]}]}]},endpoints:[`speech_generation`],snapshots:[`tts-1-hd`],family:`tts`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/tts-1-hd`,open_weight:!1},{id:`tts-1`,name:`TTS-1`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,description:`TTS is a model that converts text to natural sounding spoken text. The tts-1 model is optimized for realtime text-to-speech use cases. Use it with the Speech endpoint in the Audio API.`,status:`active`,model_type:`tts`,performance:2,speed:4,tagline:`Text-to-speech model optimized for speed`,modalities:{input:[`text`],output:[`audio`]},pricing:{tiers:[{label:`Pricing`,unit:`Per 1M tokens`,columns:[`Cost`],rows:[{label:`Standard`,values:[15]}]}]},endpoints:[`speech_generation`],snapshots:[`tts-1`],family:`tts`,license:`proprietary`,page_url:`https://platform.openai.com/docs/models/tts-1`,open_weight:!1},{id:`whisper-1`,name:`Whisper`,created_by:`openai`,source:`official`,last_updated:`2026-03-23`,family:`whisper`,description:`Whisper is a general-purpose speech recognition model, trained on a large dataset of diverse audio. You can also use it as a multitask model to perform multilingual speech recognition as well as speech translation and language identification.`,status:`active`,model_type:`transcription`,performance:2,speed:3,tagline:`General-purpose speech recognition model`,modalities:{input:[`audio`],output:[`text`]},pricing:{tiers:[{label:`Pricing`,unit:`Per 1M tokens`,columns:[`Cost`],rows:[{label:`Standard`,values:[.006]}]}]},endpoints:[`transcription`,`translation`],snapshots:[`whisper-1`],license:`proprietary`,page_url:`https://platform.openai.com/docs/models/whisper-1`,open_weight:!1}]},t=e.models.map(t=>({...t,provider:e.id}));exports.models=t,exports.provider=e;