crfm-helm 0.5.6__py3-none-any.whl → 0.5.10__py3-none-any.whl

helm/config/model_metadata.yaml

@@ -278,7 +278,7 @@ models:
   # https://aws.amazon.com/ai/generative-ai/nova/
   - name: amazon/nova-premier-v1:0
     display_name: Amazon Nova Premier
-    description: Amazon Nova Premier is the most capable model in the Nova family of foundation models. ([blog](https://aws.amazon.com/blogs/aws/amazon-nova-premier-our-most-capable-model-for-complex-tasks-and-teacher-for-model-distillation/))
+    description: Amazon Nova Premier is a capable multimodal foundation model and teacher for model distillation that processes text, images, and videos with a one-million token context window. ([model card](https://www.amazon.science/publications/amazon-nova-premier-technical-report-and-model-card), [blog](https://aws.amazon.com/blogs/aws/amazon-nova-premier-our-most-capable-model-for-complex-tasks-and-teacher-for-model-distillation/))
     creator_organization_name: Amazon
     access: limited
     release_date: 2025-04-30
@@ -286,7 +286,7 @@ models:
 
   - name: amazon/nova-pro-v1:0
     display_name: Amazon Nova Pro
-    description: Amazon Nova Pro Model
+    description: Amazon Nova Pro is a highly capable multimodal model that balances of accuracy, speed, and cost for a wide range of tasks ([model card](https://www.amazon.science/publications/the-amazon-nova-family-of-models-technical-report-and-model-card))
     creator_organization_name: Amazon
     access: limited
     release_date: 2024-12-03
@@ -294,7 +294,7 @@ models:
 
   - name: amazon/nova-lite-v1:0
     display_name: Amazon Nova Lite
-    description: Amazon Nova Lite Model
+    description: Amazon Nova Lite is a low-cost multimodal model that is fast for processing images, video, documents and text. ([model card](https://www.amazon.science/publications/the-amazon-nova-family-of-models-technical-report-and-model-card))
     creator_organization_name: Amazon
     access: limited
     release_date: 2024-12-03
@@ -302,7 +302,7 @@ models:
 
   - name: amazon/nova-micro-v1:0
     display_name: Amazon Nova Micro
-    description: Amazon Nova Micro Model
+    description: Amazon Nova Micro is a text-only model that delivers low-latency responses at low cost. ([model card](https://www.amazon.science/publications/the-amazon-nova-family-of-models-technical-report-and-model-card))
     creator_organization_name: Amazon
     access: limited
     release_date: 2024-12-03
@@ -555,6 +555,14 @@ models:
     release_date: 2025-05-14
     tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: anthropic/claude-sonnet-4-5-20250929
+    display_name: Claude 4.5 Sonnet (20250929)
+    description: Claude 4.5 Sonnet is a model from Anthropic that shows particular strengths in software coding, in agentic tasks where it runs in a loop and uses tools, and in using computers. ([blog](https://www.anthropic.com/news/claude-sonnet-4-5), [system card](https://assets.anthropic.com/m/12f214efcc2f457a/original/Claude-Sonnet-4-5-System-Card.pdf))
+    creator_organization_name: Anthropic
+    access: limited
+    release_date: 2025-09-29
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   - name: anthropic/stanford-online-all-v4-s3
     display_name: Anthropic-LM v4-s3 (52B)
     description: A 52B parameter language model, trained using reinforcement learning from human feedback [paper](https://arxiv.org/pdf/2204.05862.pdf).
@@ -946,6 +954,24 @@ models:
     release_date: 2025-01-20
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: deepseek-ai/deepseek-r1-distill-llama-70b
+    display_name: DeepSeek-R1-Distill-Llama-70B
+    description: DeepSeek-R1-Distill-Llama-70B is a fine-tuned open-source models based on Llama-3.3-70B-Instruct using samples generated by DeepSeek-R1.
+    creator_organization_name: DeepSeek
+    access: open
+    num_parameters: 70600000000
+    release_date: 2025-01-20
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: deepseek-ai/deepseek-r1-distill-qwen-14b
+    display_name: DeepSeek-R1-Distill-Qwen-14B
+    description: DeepSeek-R1-Distill-Qwen-14B is a fine-tuned open-source models based on Qwen2.5-14B using samples generated by DeepSeek-R1.
+    creator_organization_name: DeepSeek
+    access: open
+    num_parameters: 14800000000
+    release_date: 2025-01-20
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   - name: deepseek-ai/deepseek-coder-6.7b-instruct
     display_name: DeepSeek-Coder-6.7b-Instruct
     description: DeepSeek-Coder-6.7b-Instruct is a model that is fine-tuned from the LLaMA 6.7B model for the DeepSeek-Coder task.
@@ -1207,7 +1233,7 @@ models:
 
   - name: google/gemini-2.0-flash-001
     display_name: Gemini 2.0 Flash
-    description: Gemini 2.0 Flash ([documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
+    description: Gemini 2.0 Flash is a member of the Gemini 2.0 series of models, a suite of highly-capable, natively multimodal models designed to power agentic systems. ([model card](https://storage.googleapis.com/model-cards/documents/gemini-2-flash.pdf), [documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
     creator_organization_name: Google
     access: limited
     release_date: 2025-02-01
@@ -1215,7 +1241,7 @@ models:
 
   - name: google/gemini-2.0-flash-lite-preview-02-05
     display_name: Gemini 2.0 Flash Lite (02-05 preview)
-    description: Gemini 2.0 Flash Lite (02-05 preview) ([documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
+    description: Gemini 2.0 Flash Lite (02-05 preview) ([model card](https://storage.googleapis.com/model-cards/documents/gemini-2-flash.pdf), [documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
     creator_organization_name: Google
     access: limited
     release_date: 2025-02-05
@@ -1223,7 +1249,7 @@ models:
 
   - name: google/gemini-2.0-flash-lite-001
     display_name: Gemini 2.0 Flash Lite
-    description: Gemini 2.0 Flash Lite ([documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
+    description: Gemini 2.0 Flash Lite is the fastest and most cost efficient Flash model in the Gemini 2.0 series of models, a suite of highly-capable, natively multimodal models designed to power agentic systems. ([model card](https://storage.googleapis.com/model-cards/documents/gemini-2-flash.pdf), [documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
     creator_organization_name: Google
     access: limited
     release_date: 2025-03-25
@@ -1253,6 +1279,14 @@ models:
     release_date: 2025-06-17
     tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, AUDIO_LANGUAGE_MODEL_TAG, GOOGLE_GEMINI_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: google/gemini-2.5-flash-lite
+    display_name: Gemini 2.5 Flash-Lite
+    description: Gemini 2.5 Flash-Lite ([blog](https://blog.google/products/gemini/gemini-2-5-model-family-expands/))
+    creator_organization_name: Google
+    access: limited
+    release_date: 2025-07-22
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, AUDIO_LANGUAGE_MODEL_TAG, GOOGLE_GEMINI_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   - name: google/gemini-2.5-flash-preview-04-17
     display_name: Gemini 2.5 Flash (04-17 preview)
     description: Gemini 2.5 Flash (04-17 preview) ([documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
@@ -1372,6 +1406,14 @@ models:
     access: open
     release_date: 2024-06-27
     tags: [TEXT_MODEL_TAG, GOOGLE_GEMMA_INSTRUCT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  
+  - name: google/medgemma-4b-it
+    display_name: MedGemma (4B)
+    description: Gemma is a family of lightweight, open models built from the research and technology that Google used to create the Gemini models. ([model card](https://www.kaggle.com/models/google/gemma), [blog post](https://blog.google/technology/developers/gemma-open-models/))
+    creator_organization_name: Google
+    access: open
+    release_date: 2025-05-20
+    tags: [TEXT_MODEL_TAG, GOOGLE_GEMMA_INSTRUCT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
   - name: google/paligemma-3b-mix-224
     display_name: PaliGemma (3B) Mix 224
@@ -2573,6 +2615,14 @@ models:
     release_date: 2025-05-07
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: mistralai/mistral-medium-3.1
+    display_name: Mistral Medium 3.1
+    description: Mistral Medium 3.1 is a language model that is intended to to deliver state-of-the-art performance at lower cost. ([blog](https://mistral.ai/news/mistral-medium-3))
+    creator_organization_name: Mistral AI
+    access: limited
+    release_date: 2025-05-07
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   - name: mistralai/mistral-large-2402
     display_name: Mistral Large (2402)
     description: Mistral Large is a multilingual model with a 32K tokens context window and function-calling capabilities. ([blog](https://mistral.ai/news/mistral-large/))
@@ -2624,6 +2674,15 @@ models:
     release_date: 2024-11-18
     tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  # Moonshot AI
+  - name: moonshotai/kimi-k2-instruct
+    display_name: Kimi K2 Instruct
+    description:  Kimi K2 Instruct is a mixture-of-experts (MoE) language model with 32 billion activated parameters and 1 trillion total parameters trained with the Muon optimizer on 15.5T tokens. ([blog](https://moonshotai.github.io/Kimi-K2/))
+    creator_organization_name: Moonshot AI
+    access: open
+    num_parameters: 1029173256720
+    release_date: 2024-07-14  # Blog post has no date, so use the date from this news article https://www.cnbc.com/2025/07/14/alibaba-backed-moonshot-releases-kimi-k2-ai-rivaling-chatgpt-claude.html
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
   # MosaicML
   - name: mosaicml/mpt-7b
@@ -3043,6 +3102,30 @@ models:
     release_date: 2025-04-14
     tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: openai/gpt-5-2025-08-07
+    display_name: GPT-5 (2025-08-07)
+    description: GPT-5 (2025-08-07) is a multimdodal model trained for real-world coding tasks and long-running agentic tasks. ([blog](https://openai.com/index/introducing-gpt-5-for-developers/), [system card](https://cdn.openai.com/pdf/8124a3ce-ab78-4f06-96eb-49ea29ffb52f/gpt5-system-card-aug7.pdf))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-08-07
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/gpt-5-mini-2025-08-07
+    display_name: GPT-5 mini (2025-08-07)
+    description: GPT-5 mini (2025-08-07) is a multimdodal model trained for real-world coding tasks and long-running agentic tasks. ([blog](https://openai.com/index/introducing-gpt-5-for-developers/), [system card](https://cdn.openai.com/pdf/8124a3ce-ab78-4f06-96eb-49ea29ffb52f/gpt5-system-card-aug7.pdf))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-08-07
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/gpt-5-nano-2025-08-07
+    display_name: GPT-5 nano (2025-08-07)
+    description: GPT-5 nano (2025-08-07) is a multimdodal model trained for real-world coding tasks and long-running agentic tasks. ([blog](https://openai.com/index/introducing-gpt-5-for-developers/), [system card](https://cdn.openai.com/pdf/8124a3ce-ab78-4f06-96eb-49ea29ffb52f/gpt5-system-card-aug7.pdf))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-08-07
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   - name: openai/whisper-1_gpt-4o-2024-11-20
     display_name: Whisper-1 + GPT-4o (2024-11-20)
     description: Transcribes the text with Whisper-1 and then uses GPT-4o to generate a response.
@@ -3256,6 +3339,31 @@ models:
     release_date: 2025-04-16
     tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: openai/o3-pro-2025-06-10-high-reasoning-effort
+    display_name: o3-pro (2025-06-10, high reasoning effort)
+    description: o3-pro is an o-series model designed to think longer and provide the most reliable responses. ([blog post](https://help.openai.com/en/articles/9624314-model-release-notes))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-06-10
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  ## GPT-OSS
+  - name: openai/gpt-oss-20b
+    display_name: gpt-oss-20b
+    description: gpt-oss-20b is an open-weight language model that was trained using a mix of reinforcement learning and other techniques informed by OpenAI's internal models. It uses a mixture-of-experts architecture and activates 3.6B parameters per token. ([blog](https://openai.com/index/introducing-gpt-oss/))
+    creator_organization_name: OpenAI
+    access: open
+    release_date: 2025-08-05
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/gpt-oss-120b
+    display_name: gpt-oss-120b
+    description: gpt-oss-120b is an open-weight language model that was trained using a mix of reinforcement learning and other techniques informed by OpenAI's internal models. It uses a mixture-of-experts architecture and activates 5.1B parameters per token. ([blog](https://openai.com/index/introducing-gpt-oss/))
+    creator_organization_name: OpenAI
+    access: open
+    release_date: 2025-08-05
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   ## Codex Models
   # DEPRECATED: Codex models have been shut down on March 23 2023.
 
@@ -3532,6 +3640,22 @@ models:
     release_date: 2025-04-29
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: qwen/qwen3-next-80b-a3b-thinking
+    display_name: Qwen3-Next 80B A3B Thinking
+    description: Qwen3-Next is a new model architecture for improving training and inference efficiency under long-context and large-parameter settings. Compared to the MoE structure of Qwen3, Qwen3-Next introduces a hybrid attention mechanism, a highly sparse Mixture-of-Experts (MoE) structure, training-stability-friendly optimizations, and a multi-token prediction mechanism for faster inference. ([blog](https://qwen.ai/blog?id=4074cca80393150c248e508aa62983f9cb7d27cd&from=research.latest-advancements-list))
+    creator_organization_name: Qwen
+    access: open
+    release_date: 2025-07-21  # https://x.com/Alibaba_Qwen/status/1947344511988076547
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: qwen/qwen3-235b-a22b-instruct-2507-fp8
+    display_name: Qwen3 235B A22B Instruct 2507 FP8
+    description: Qwen3 235B A22B Instruct 2507 FP8 is an updated version of the non-thinking mode of Qwen3 235B A22B FP8.
+    creator_organization_name: Qwen
+    access: open
+    release_date: 2025-07-21  # https://x.com/Alibaba_Qwen/status/1947344511988076547
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   - name: qwen/qwq-32b-preview
     display_name: QwQ (32B Preview)
     description: QwQ-32B-Preview is an experimental research model developed by the Qwen Team, focused on advancing AI reasoning capabilities. ([blog post](https://qwenlm.github.io/blog/qwq-32b-preview/)).
@@ -3875,7 +3999,190 @@ models:
     release_date: 2023-05-25
     tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG]
 
+  - name: tiiuae/falcon3-1b-instruct
+    display_name: Falcon3-1B-Instruct
+    description: Falcon3-1B-Instruct is an open-weights foundation model that supports 4 languages (English, French, Spanish, Portuguese) that was trained on 14T tokens.
+    creator_organization_name: TII UAE
+    access: open
+    num_parameters: 1670000000
+    release_date: 2024-12-17  # https://huggingface.co/docs/transformers/main/en/model_doc/falcon3
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: tiiuae/falcon3-3b-instruct
+    display_name: Falcon3-3B-Instruct
+    description: Falcon3-3B-Instruct is an open-weights foundation model that supports 4 languages (English, French, Spanish, Portuguese) that was trained on 14T tokens.
+    creator_organization_name: TII UAE
+    access: open
+    num_parameters: 3230000000
+    release_date: 2024-12-17  # https://huggingface.co/docs/transformers/main/en/model_doc/falcon3
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: tiiuae/falcon3-7b-instruct
+    display_name: Falcon3-7B-Instruct
+    description: Falcon3-7B-Instruct is an open-weights foundation model that supports 4 languages (English, French, Spanish, Portuguese) that was trained on 14T tokens.
+    creator_organization_name: TII UAE
+    access: open
+    num_parameters: 7460000000
+    release_date: 2024-12-17  # https://huggingface.co/docs/transformers/main/en/model_doc/falcon3
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: tiiuae/falcon3-10b-instruct
+    display_name: Falcon3-10B-Instruct
+    description: Falcon3-10B-Instruct is an open-weights foundation model that supports 4 languages (English, French, Spanish, Portuguese) that was trained on 14T tokens.
+    creator_organization_name: TII UAE
+    access: open
+    num_parameters: 10300000000
+    release_date: 2024-12-17  # https://huggingface.co/docs/transformers/main/en/model_doc/falcon3
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  # AceGPT-v2
+  - name: freedomintelligence/acegpt-v2-8b-chat
+    display_name: AceGPT-v2-8B-Chat
+    description: AceGPT is a fully fine-tuned generative text model collection, particularly focused on the Arabic language domain. AceGPT-v2-8B-Chat is based on Meta-Llama-3-8B. ([paper](https://arxiv.org/abs/2412.12310))
+    creator_organization_name: FreedomAI
+    access: open
+    num_parameters: 8030000000
+    release_date: 2024-10-20
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: freedomintelligence/acegpt-v2-32b-chat
+    display_name: AceGPT-v2-32B-Chat
+    description: AceGPT is a fully fine-tuned generative text model collection, particularly focused on the Arabic language domain. AceGPT-v2-32B-Chat is based on Qwen1.5-32B. ([paper](https://arxiv.org/abs/2412.12310))
+    creator_organization_name: FreedomAI
+    access: open
+    num_parameters: 32500000000
+    release_date: 2024-10-20
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: freedomintelligence/acegpt-v2-70b-chat
+    display_name: AceGPT-v2-70B-Chat
+    description: AceGPT is a fully fine-tuned generative text model collection, particularly focused on the Arabic language domain. AceGPT-v2-70B-Chat is based on Meta-Llama-3-70B. ([paper](https://arxiv.org/abs/2412.12310))
+    creator_organization_name: FreedomAI
+    access: open
+    num_parameters: 70600000000
+    release_date: 2024-10-20
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  # ALLaM
+  - name: allam-ai/allam-7b-instruct-preview
+    display_name: ALLaM-7B-Instruct-preview
+    description: ALLaM-7B-Instruct-preview is a model designed to advance Arabic language technology, which used a recipe of training on 4T English tokens followed by training on 1.2T mixed Arabic/English tokens. ([paper](https://arxiv.org/abs/2407.15390v1))
+    creator_organization_name: NCAI & SDAIA
+    access: open
+    num_parameters: 7000000000
+    release_date: 2024-07-22
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  # SILMA
+  - name: silma-ai/silma-9b-instruct-v1.0
+    display_name: SILMA 9B
+    description: SILMA 9B is a compact Arabic language model based on Google Gemma. ([model card](https://huggingface.co/silma-ai/SILMA-9B-Instruct-v1.0))
+    creator_organization_name: SILMA AI
+    access: open
+    num_parameters: 9240000000
+    release_date: 2024-08-17
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  # Jais Family
+
+  - name: inceptionai/jais-family-590m-chat
+    display_name: Jais-family-590m-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 771000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-family-1p3b-chat
+    display_name: Jais-family-1p3b-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 1560000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-family-2p7b-chat
+    display_name: Jais-family-2p7b-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 2950000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-family-6p7b-chat
+    display_name: Jais-family-6p7b-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 7140000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-family-6p7b-chat
+    display_name: Jais-family-6p7b-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 7140000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-family-13b-chat
+    display_name: Jais-family-13b-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 13500000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-family-30b-8k-chat
+    display_name: Jais-family-30b-8k-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 30800000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-family-30b-16k-chat
+    display_name: Jais-family-30b-16k-chat
+    description: The Jais family of models is a series of bilingual English-Arabic large language models (LLMs) that are trained from scratch and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 30800000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-adapted-7b-chat
+    display_name: Jais-adapted-7b-chat
+    description: The Jais adapted models are bilingual English-Arabic large language models (LLMs) that are trained adaptively from Llama-2 and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 7000000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-adapted-13b-chat
+    display_name: Jais-adapted-13b-chat
+    description: The Jais adapted models are bilingual English-Arabic large language models (LLMs) that are trained adaptively from Llama-2 and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 13300000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: inceptionai/jais-adapted-70b-chat
+    display_name: Jais-adapted-70b-chat
+    description: The Jais adapted models are bilingual English-Arabic large language models (LLMs) that are trained adaptively from Llama-2 and optimized to excel in Arabic while having strong English capabilities. ([website](https://inceptionai.ai/jaisfamily/index.html), [blog](https://mbzuai.ac.ae/news/meet-jais-the-worlds-most-advanced-arabic-large-language-model-open-sourced-by-g42s-inception/))
+    creator_organization_name: Inception
+    access: open
+    num_parameters: 69500000000
+    release_date: 2023-08-30
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
   # Together
   - name: together/gpt-jt-6b-v1
@@ -4108,7 +4415,15 @@ models:
     description: Palmyra X5 is a language model for enterprise that uses a Mixture of Experts (MoE) architecture and a hybrid attention mechanism that blends linear and softmax attention. ([blog](https://writer.com/engineering/long-context-palmyra-x5/))
     creator_organization_name: Writer
     access: limited
-    release_date: 2024-04-28
+    release_date: 2025-04-28
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: writer/palmyra-x5-v1-bedrock
+    display_name: Palmyra X5 (Bedrock)
+    description: Palmyra X5 is a language model for enterprise that uses a Mixture of Experts (MoE) architecture and a hybrid attention mechanism that blends linear and softmax attention. ([blog](https://writer.com/engineering/long-context-palmyra-x5/)) This is the model verison that is hosted on Bedrock. ([blog](https://aws.amazon.com/blogs/aws/writer-palmyra-x5-and-x4-foundation-models-are-now-available-in-amazon-bedrock/))
+    creator_organization_name: Writer
+    access: limited
+    release_date: 2025-04-28
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
   - name: writer/palmyra-med-32k
@@ -4163,6 +4478,14 @@ models:
     release_date: 2025-04-03  # https://docs.x.ai/docs/release-notes#april-2025
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: xai/grok-4-0709
+    display_name: Grok 4 (0709)
+    description: Grok 4 (0709) is a model that includes native tool use and real-time search integration. ([blog](https://x.ai/news/grok-4))
+    creator_organization_name: xAI
+    access: limited
+    release_date: 2025-07-09
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
   # Yandex
   - name: yandex/yalm
     display_name: YaLM (100B)
@@ -4266,6 +4589,42 @@ models:
     release_date: 2023-11-08
     tags: [TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
+  - name: maritaca-ai/sabiazinho-3
+    display_name: Sabiazinho 3
+    description: Sabiazinho-3 is a decoder-only language model designed for Portuguese text generation and understanding tasks. It supports a long context window of up to 128,000 tokens and is offered via API with scalable rate limits. The model is trained on diverse Portuguese corpora with knowledge up to july 2023.
+    creator_organization_name: Maritaca AI
+    access: limited
+    release_date: 2025-02-06
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: maritaca-ai/sabia-3
+    display_name: Sabía 3
+    description: Sabiá-3 is a decoder-only language model designed for Portuguese text generation and understanding tasks. It supports a long context window of up to 128,000 tokens and is offered via API with scalable rate limits. The model is trained on diverse Portuguese corpora with knowledge up to july 2023.
+    creator_organization_name: Maritaca AI
+    access: limited
+    release_date: 2024-12-11
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: maritaca-ai/sabia-3.1-2025-05-08
+    display_name: Sabía 3.1
+    description: Sabiá-3.1 is a decoder-only language model designed for Portuguese text generation and understanding tasks. It supports a long context window of up to 128,000 tokens and is offered via API with scalable rate limits. The model is trained on diverse Portuguese corpora with knowledge up to August 2024.
+    creator_organization_name: Maritaca AI
+    access: limited
+    release_date: 2025-05-08
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG] 
+
+  # Z.ai
+
+  - name: zai-org/glm-4.5-air-fp8
+    display_name: GLM-4.5-Air-FP8
+    description: GLM-4.5-Air-FP8 is a hybrid reasoning model designed to unify reasoning, coding, and agentic capabilities into a single model. It has 106 billion total parameters and 12 billion active parameters. The thinking mode is enabled by default. ([blog](https://z.ai/blog/glm-4.5))
+    creator_organization_name: Z.ai
+    access: open
+    num_parameters: 110000000000
+    release_date: 2025-07-28
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+
 # Granite - IBM
 # https://www.ibm.com/granite
 # https://github.com/ibm-granite/granite-3.0-language-models
@@ -4479,21 +4838,61 @@ models:
     tags: [ TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG ]
 
   - name: ibm/granite-3.3-8b-instruct
-    display_name: Granite 3.3 8B Instruct
-    description: Granite 3.3 8B Instruct is a 8-billion parameter 128K context length language model fine-tuned for improved reasoning and instruction-following capabilities. ([model card](https://huggingface.co/ibm-granite/granite-3.3-8b-instruct))
+    display_name: IBM Granite 3.3 8B Instruct
+    description: IBM Granite 3.3 8B Instruct is an 8-billion parameter 128K context length language model fine-tuned for improved reasoning and instruction-following capabilities. ([model card](https://huggingface.co/ibm-granite/granite-3.3-8b-instruct))
     creator_organization_name: IBM
     access: open
     num_parameters: 8170000000
     release_date: 2025-04-16
     tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, ABLATION_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
-  - name: mistralai/mixtral-8x7b-instruct-v0:1
-    display_name: Mixtral 8x7B Instruct on IBM WatsonX
-    description: A 7B sparse Mixture-of-Experts model with stronger capabilities than Mistral 7B. Uses 12B active parameters out of 45B total. Supports multiple languages, code and 32k context window.
-    creator_organization_name: Mistral
-    access: limited
-    release_date: 2023-12-11
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, ABLATION_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+  - name: ibm/granite-3.3-8b-instruct-with-guardian
+    display_name: IBM Granite 3.3 8B Instruct (with guardian)
+    description: IBM Granite 3.3 8B Instruct is an 8-billion parameter 128K context length language model fine-tuned for improved reasoning and instruction-following capabilities. All prompts were first evaluated for risk by [IBM Granite Guardian 3.2 5B](https://www.ibm.com/granite/docs/models/guardian/) and prompts that were deemed risky (with a risk threshold of 0.8) received the response "I'm very sorry, but I can't assist with that.". ([model card](https://huggingface.co/ibm-granite/granite-3.3-8b-instruct))
+    creator_organization_name: IBM
+    access: open
+    num_parameters: 8170000000
+    release_date: 2025-04-16
+    # Unfortunately this setup used a IBM internal API endpoint that is not publicly available, so we mark it with DEPRECATED_MODEL_TAG
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, ABLATION_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: ibm/granite-4.0-h-small
+    display_name: IBM Granite 4.0 Small
+    description: IBM Granite 4.0 Small is a hybrid model with 32B total parameters and 9B active parameters that uses the Mixture of Experts (MoE) routing strategy with Mamba-2 and Transformer-based self-attention components.
+    creator_organization_name: IBM
+    access: open
+    num_parameters: 32200000000
+    release: 2025-10-02
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: ibm/granite-4.0-micro
+    display_name: IBM Granite 4.0 Micro
+    description: IBM Granite 4.0 Micro is a dense Transformer model with 3B total parameters that provides an alternative option for users when Mamba2 support is not yet optimized.
+    creator_organization_name: IBM
+    access: open
+    num_parameters: 3400000000
+    release: 2025-10-02
+    tags: [TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: ibm/granite-4.0-h-small-with-guardian
+    display_name: IBM Granite 4.0 Small (with guardian)
+    description: IBM Granite 4.0 Small is a hybrid model with 32B total parameters and 9B active parameters that uses the Mixture of Experts (MoE) routing strategy with Mamba-2 and Transformer-based self-attention components.
+    creator_organization_name: IBM
+    access: open
+    num_parameters: 32200000000
+    release: 2025-10-02
+    # Unfortunately this setup used a IBM internal API endpoint that is not publicly available, so we mark it with DEPRECATED_MODEL_TAG
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: ibm/granite-4.0-micro-with-guardian
+    display_name: IBM Granite 4.0 Micro (with guardian)
+    description: IBM Granite 4.0 Micro is a dense Transformer model with 3B total parameters that provides an alternative option for users when Mamba2 support is not yet optimized.
+    creator_organization_name: IBM
+    access: open
+    num_parameters: 3400000000
+    release: 2025-10-02
+    # Unfortunately this setup used a IBM internal API endpoint that is not publicly available, so we mark it with DEPRECATED_MODEL_TAG
+    tags: [DEPRECATED_MODEL_TAG, TEXT_MODEL_TAG, FULL_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
 
   - name: ura-hcmut/ura-llama-2.1-8b
     display_name: URA-Llama 2.1 (8B)
@@ -4682,4 +5081,189 @@ models:
     access: open
     num_parameters: 4000000000
     release_date: 2024-04-02
-    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: CEIA-UFG/Gemma-3-Gaia-PT-BR-4b-it
+    display_name: Gemma-3 Gaia PT-BR 4b Instruct
+    description: Gemma-3 Gaia PT-BR 4b Instruct is a model trained by CEIA-UFG for understanding and generating Brazilian Portuguese text.
+    creator_organization_name: CEIA-UFG
+    access: open
+    num_parameters: 4000000000
+    release_date: 2025-06-01
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: recogna-nlp/bode-13b-alpaca-pt-br-no-peft
+    display_name: Bode 13B Alpaca PT-BR
+    description: Bode is a language model (LLM) for Portuguese, based on LLaMA 2 and fine-tuned with the Alpaca dataset translated into Portuguese. Suitable for instruction, text generation, translation and tasks in Portuguese.
+    creator_organization_name: Recogna NLP
+    access: open
+    num_parameters: 13000000000
+    release_date: 2024-01-05
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: 22h/cabrita_7b_pt_850000
+    display_name: Cabrita PT-BR 7B
+    description: Cabrita is an OpenLLaMA-based model, continuously trained in Portuguese (mC4-pt subset) for 850000 steps with efficient tokenization adapted to the language.
+    creator_organization_name: 22h
+    access: open
+    num_parameters: 7000000000
+    release_date: 2023-08-23
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: PORTULAN/gervasio-7b-portuguese-ptbr-decoder
+    display_name: Gervásio PT-BR/PT-PT 7B Decoder
+    description: Gervásio PT* is a 7B parameter decoder model, adapted from LLaMA27B, trained for both Brazilian and European Portuguese. Fine-tuned with translated data from benchmarks such as GLUE and SuperGLUE.
+    creator_organization_name: PORTULAN (University of Lisbon NLX)
+    access: open
+    num_parameters: 6740000000
+    release_date: 2024-02-29
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: TucanoBR/Tucano-2b4
+    display_name: Tucano PT-BR 2b4
+    description: Tucano is a series of decoder models based on LLaMA2, natively pre-trained in Portuguese using the GigaVerbo dataset (200B tokens), with the 2B model trained for 1.96M steps over 845h (515B tokens, 4 epochs).
+    creator_organization_name: TucanoBR (University of Bonn)
+    access: open
+    num_parameters: 2444618240
+    release_date: 2024-12-11
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: nicholasKluge/TeenyTinyLlama-460m
+    display_name: TeenyTinyLlama 460M PT-BR
+    description: TeenyTinyLlama-460m is a lightweight and efficient model based on LLaMA2, trained exclusively on Brazilian Portuguese. It uses RoPE embeddings and SwiGLU activations, with a refined SentencePiece tokenizer and a low-resource optimized architecture.
+    creator_organization_name: Nicholas Kluge.
+    access: open
+    num_parameters: 460000000
+    release_date: 2024-01-30
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  # DSPy Models (EXPERIMENTAL)
+  # The following model configurations use the DSPyClient for inference with DSPy modules.
+  
+  - name: anthropic/claude-3-7-sonnet-20250219-dspy-zs-predict
+    display_name: Claude 3.7 Sonnet (20250219) (DSPy Zero-Shot Predict)
+    description: Claude 3.7 Sonnet is a Claude 3 family hybrid reasoning model that can produce near-instant responses or extended, step-by-step thinking that is made visible to the user ([blog](https://www.anthropic.com/news/claude-3-7-sonnet)).
+    creator_organization_name: Anthropic
+    access: limited
+    release_date: 2025-02-24
+    tags: [ANTHROPIC_CLAUDE_3_MODEL_TAG, TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: google/gemini-2.0-flash-001-dspy-zs-predict
+    display_name: Gemini 2.0 Flash (DSPy Zero-Shot Predict)
+    description: Gemini 2.0 Flash is a member of the Gemini 2.0 series of models, a suite of highly-capable, natively multimodal models designed to power agentic systems. ([model card](https://storage.googleapis.com/model-cards/documents/gemini-2-flash.pdf), [documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
+    creator_organization_name: Google
+    access: limited
+    release_date: 2025-02-01
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, AUDIO_LANGUAGE_MODEL_TAG, GOOGLE_GEMINI_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/gpt-4o-2024-05-13-dspy-zs-predict
+    display_name: GPT-4o (2024-05-13) (DSPy Zero-Shot Predict)
+    description: GPT-4o (2024-05-13) is a large multimodal model that accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. ([blog](https://openai.com/index/hello-gpt-4o/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2024-04-09
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/o3-mini-2025-01-31-dspy-zs-predict
+    display_name: o3-mini (2025-01-31) (DSPy Zero-Shot Predict)
+    description: o3-mini is a small reasoning model form OpenAI that aims to deliver STEM capabilities while maintaining the low cost and reduced latency of OpenAI o1-mini. ([blog post](https://openai.com/index/openai-o3-mini/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-01-31
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: anthropic/claude-3-7-sonnet-20250219-dspy-zs-cot
+    display_name: Claude 3.7 Sonnet (20250219) (DSPy Zero-Shot ChainOfThought)
+    description: Claude 3.7 Sonnet is a Claude 3 family hybrid reasoning model that can produce near-instant responses or extended, step-by-step thinking that is made visible to the user ([blog](https://www.anthropic.com/news/claude-3-7-sonnet)).
+    creator_organization_name: Anthropic
+    access: limited
+    release_date: 2025-02-24
+    tags: [ANTHROPIC_CLAUDE_3_MODEL_TAG, TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: google/gemini-2.0-flash-001-dspy-zs-cot
+    display_name: Gemini 2.0 Flash (DSPy Zero-Shot ChainOfThought)
+    description: Gemini 2.0 Flash is a member of the Gemini 2.0 series of models, a suite of highly-capable, natively multimodal models designed to power agentic systems. ([model card](https://storage.googleapis.com/model-cards/documents/gemini-2-flash.pdf), [documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
+    creator_organization_name: Google
+    access: limited
+    release_date: 2025-02-01
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, AUDIO_LANGUAGE_MODEL_TAG, GOOGLE_GEMINI_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/gpt-4o-2024-05-13-dspy-zs-cot
+    display_name: GPT-4o (2024-05-13) (DSPy Zero-Shot ChainOfThought)
+    description: GPT-4o (2024-05-13) is a large multimodal model that accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. ([blog](https://openai.com/index/hello-gpt-4o/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2024-04-09
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/o3-mini-2025-01-31-dspy-zs-cot
+    display_name: o3-mini (2025-01-31) (DSPy Zero-Shot ChainOfThought)
+    description: o3-mini is a small reasoning model form OpenAI that aims to deliver STEM capabilities while maintaining the low cost and reduced latency of OpenAI o1-mini. ([blog post](https://openai.com/index/openai-o3-mini/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-01-31
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: anthropic/claude-3-7-sonnet-20250219-dspy-fs-bfrs
+    display_name: Claude 3.7 Sonnet (20250219) (DSPy BootstrapFewShotWithRandomSearch)
+    description: Claude 3.7 Sonnet is a Claude 3 family hybrid reasoning model that can produce near-instant responses or extended, step-by-step thinking that is made visible to the user ([blog](https://www.anthropic.com/news/claude-3-7-sonnet)).
+    creator_organization_name: Anthropic
+    access: limited
+    release_date: 2025-02-24
+    tags: [ANTHROPIC_CLAUDE_3_MODEL_TAG, TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: google/gemini-2.0-flash-001-dspy-fs-bfrs
+    display_name: Gemini 2.0 Flash (DSPy BootstrapFewShotWithRandomSearch)
+    description: Gemini 2.0 Flash is a member of the Gemini 2.0 series of models, a suite of highly-capable, natively multimodal models designed to power agentic systems. ([model card](https://storage.googleapis.com/model-cards/documents/gemini-2-flash.pdf), [documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
+    creator_organization_name: Google
+    access: limited
+    release_date: 2025-02-01
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, AUDIO_LANGUAGE_MODEL_TAG, GOOGLE_GEMINI_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/gpt-4o-2024-05-13-dspy-fs-bfrs
+    display_name: GPT-4o (2024-05-13) (DSPy BootstrapFewShotWithRandomSearch)
+    description: GPT-4o (2024-05-13) is a large multimodal model that accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. ([blog](https://openai.com/index/hello-gpt-4o/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2024-04-09
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/o3-mini-2025-01-31-dspy-fs-bfrs
+    display_name: o3-mini (2025-01-31) (DSPy BootstrapFewShotWithRandomSearch)
+    description: o3-mini is a small reasoning model form OpenAI that aims to deliver STEM capabilities while maintaining the low cost and reduced latency of OpenAI o1-mini. ([blog post](https://openai.com/index/openai-o3-mini/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-01-31
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: anthropic/claude-3-7-sonnet-20250219-dspy-fs-miprov2
+    display_name: Claude 3.7 Sonnet (20250219) (DSPy MIPROv2)
+    description: Claude 3.7 Sonnet is a Claude 3 family hybrid reasoning model that can produce near-instant responses or extended, step-by-step thinking that is made visible to the user ([blog](https://www.anthropic.com/news/claude-3-7-sonnet)).
+    creator_organization_name: Anthropic
+    access: limited
+    release_date: 2025-02-24
+    tags: [ANTHROPIC_CLAUDE_3_MODEL_TAG, TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: google/gemini-2.0-flash-001-dspy-fs-miprov2
+    display_name: Gemini 2.0 Flash (DSPy MIPROv2)
+    description: Gemini 2.0 Flash is a member of the Gemini 2.0 series of models, a suite of highly-capable, natively multimodal models designed to power agentic systems. ([model card](https://storage.googleapis.com/model-cards/documents/gemini-2-flash.pdf), [documentation](https://ai.google.dev/gemini-api/docs/models/gemini))
+    creator_organization_name: Google
+    access: limited
+    release_date: 2025-02-01
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, AUDIO_LANGUAGE_MODEL_TAG, GOOGLE_GEMINI_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/gpt-4o-2024-05-13-dspy-fs-miprov2
+    display_name: GPT-4o (2024-05-13) (DSPy MIPROv2)
+    description: GPT-4o (2024-05-13) is a large multimodal model that accepts as input any combination of text, audio, and image and generates any combination of text, audio, and image outputs. ([blog](https://openai.com/index/hello-gpt-4o/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2024-04-09
+    tags: [TEXT_MODEL_TAG, VISION_LANGUAGE_MODEL_TAG, OPENAI_CHATGPT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]
+
+  - name: openai/o3-mini-2025-01-31-dspy-fs-miprov2
+    display_name: o3-mini (2025-01-31) (DSPy MIPROv2)
+    description: o3-mini is a small reasoning model form OpenAI that aims to deliver STEM capabilities while maintaining the low cost and reduced latency of OpenAI o1-mini. ([blog post](https://openai.com/index/openai-o3-mini/))
+    creator_organization_name: OpenAI
+    access: limited
+    release_date: 2025-01-31
+    tags: [TEXT_MODEL_TAG, LIMITED_FUNCTIONALITY_TEXT_MODEL_TAG, INSTRUCTION_FOLLOWING_MODEL_TAG]