ai 7.0.113 → 7.0.116

package/docs/03-ai-sdk-core/15-tools-and-tool-calling.mdx

@@ -1386,12 +1386,17 @@ orders the remaining tools.
   and Google (Gemini 3 models).
 </Note>
 
-For Google, use base64 inline-data file parts
+For Google Generative AI, use base64 inline-data file parts
 (`{ type: 'file', mediaType, data: { type: 'data', data } }`) or base64
 `data:` URLs in URL-style file parts
 (`{ type: 'file', mediaType, data: { type: 'url', url: new URL('data:...') } }`).
 Remote HTTP(S) URLs in tool-result URL parts are not supported.
 
+The [Google Vertex provider](/providers/ai-sdk-providers/google-vertex#tool-result-files)
+also supports direct `gs://` tool-result file references for Gemini 3 and later
+models with supported image and document MIME types. It downloads HTTP(S)
+tool-result URLs and sends their contents as inline data.
+
 In order to send multi-modal tool results, e.g. screenshots, back to the model,
 they need to be converted into a specific format.
 

package/docs/03-ai-sdk-core/17-mcp-apps.mdx

@@ -68,7 +68,7 @@ export async function POST(req: Request) {
     const tools = client.toolsFromDefinitions(modelVisible);
 
     const result = streamText({
-      model: openai('gpt-4o-mini'),
+      model: openai('gpt-6-luna'),
       tools,
       messages: await convertToModelMessages(messages),
       onEnd: async () => {

package/docs/03-ai-sdk-core/20-prompt-engineering.mdx

@@ -13,7 +13,7 @@ When you create prompts that include tools, getting good results can be tricky a
 
 Here are a few tips to help you get the best results:
 
-1. Use a model that is strong at tool calling, such as `gpt-5` or `gpt-4.1`. Weaker models will often struggle to call tools effectively and flawlessly.
+1. Use a model that is strong at tool calling, such as `gpt-6-astra` or `claude-sonnet-5`. Weaker models will often struggle to call tools effectively and flawlessly.
 1. Keep the number of tools low, e.g. to 5 or less.
 1. Keep the complexity of the tool parameters low. Complex Zod schemas with many nested and optional elements, unions, etc. can be challenging for the model to work with.
 1. Use semantically meaningful names for your tools, parameters, parameter properties, etc. The more information you pass to the model, the better it can understand what you want.

package/docs/03-ai-sdk-core/26-reasoning.mdx

@@ -13,7 +13,7 @@ Many language models support an internal "reasoning" phase (sometimes also calle
 import { generateText } from 'ai';
 
 const { text, reasoning, reasoningText } = await generateText({
-  model: 'anthropic/claude-sonnet-4.6',
+  model: 'anthropic/claude-sonnet-5',
   reasoning: 'medium',
   prompt: 'How many people will live in the world in 2040?',
 });
@@ -39,7 +39,7 @@ The `reasoning` parameter works the same way with `streamText`:
 import { streamText } from 'ai';
 
 const result = streamText({
-  model: 'google/gemini-3-flash-preview',
+  model: 'google/gemini-3.8-flash',
   reasoning: 'high',
   prompt: 'Explain the Riemann hypothesis in simple terms.',
 });
@@ -62,7 +62,7 @@ import { generateText } from 'ai';
 import { openai } from '@ai-sdk/openai';
 
 const { text } = await generateText({
-  model: openai.responses('gpt-5.4'),
+  model: openai.responses('gpt-6-astra'),
   reasoning: 'low', // ignored because providerOptions.openai.reasoningEffort is set
   providerOptions: {
     openai: {
@@ -89,10 +89,11 @@ If you currently control reasoning via `providerOptions`, you can migrate to the
 
 ```ts
 const { text } = await generateText({
-  model: anthropic('claude-opus-4.6'),
+  model: anthropic('claude-opus-5-5'),
   providerOptions: {
     anthropic: {
-      thinking: { type: 'adaptive', effort: 'high' },
+      thinking: { type: 'adaptive' },
+      effort: 'high',
     },
   },
   prompt: 'How many people will live in the world in 2040?',
@@ -103,7 +104,7 @@ const { text } = await generateText({
 
 ```ts
 const { text } = await generateText({
-  model: anthropic('claude-opus-4.6'),
+  model: anthropic('claude-opus-5-5'),
   reasoning: 'high',
   prompt: 'How many people will live in the world in 2040?',
 });
@@ -139,10 +140,10 @@ If you need to enforce an exact token budget (e.g. exactly 12000 tokens), keep u
 
 ```ts
 const { text } = await generateText({
-  model: google('gemini-3-flash-preview'),
+  model: google('gemini-3.8-flash'),
   providerOptions: {
     google: {
-      thinkingConfig: { thinkingBudget: 4096, includeThoughts: true },
+      thinkingConfig: { thinkingLevel: 'high', includeThoughts: true },
     },
   },
   prompt: 'Explain the Riemann hypothesis in simple terms.',
@@ -153,8 +154,8 @@ const { text } = await generateText({
 
 ```ts
 const { text } = await generateText({
-  model: google('gemini-3-flash-preview'),
-  reasoning: 'medium',
+  model: google('gemini-3.8-flash'),
+  reasoning: 'high',
   providerOptions: {
     google: { thinkingConfig: { includeThoughts: true } },
   },
@@ -166,7 +167,7 @@ const { text } = await generateText({
 
 ```ts
 const { text } = await generateText({
-  model: openai.responses('o3'),
+  model: openai.responses('gpt-6-astra'),
   providerOptions: {
     openai: { reasoningEffort: 'high', reasoningSummary: 'auto' },
   },
@@ -178,7 +179,7 @@ const { text } = await generateText({
 
 ```ts
 const { text } = await generateText({
-  model: openai.responses('o3'),
+  model: openai.responses('gpt-6-astra'),
   reasoning: 'high',
   providerOptions: {
     openai: { reasoningSummary: 'auto' },

package/docs/03-ai-sdk-core/31-reranking.mdx

@@ -12,7 +12,7 @@ often producing more accurate relevance scores.
 ## Reranking Documents
 
 The AI SDK provides the [`rerank`](/docs/reference/ai-sdk-core/rerank) function to rerank documents based on their relevance to a query.
-You can use it with reranking models, e.g. `cohere.reranking('rerank-v3.5')` or `bedrock.reranking('cohere.rerank-v3-5:0')`.
+You can use it with reranking models, e.g. `cohere.reranking('rerank-v4.0-pro')` or `bedrock.reranking('cohere.rerank-v3-5:0')`.
 
 ```tsx
 import { rerank } from 'ai';
@@ -25,7 +25,7 @@ const documents = [
 ];
 
 const { ranking } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents,
   query: 'talk about rain',
   topN: 2, // Return top 2 most relevant documents
@@ -60,7 +60,7 @@ const documents = [
 ];
 
 const { ranking, rerankedDocuments } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents,
   query: 'Which pricing did we get from Oracle?',
   topN: 1,
@@ -79,7 +79,7 @@ import { cohere } from '@ai-sdk/cohere';
 import { rerank } from 'ai';
 
 const { ranking, rerankedDocuments, originalDocuments } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: ['sunny day at the beach', 'rainy afternoon in the city'],
   query: 'talk about rain',
 });
@@ -106,7 +106,7 @@ import { cohere } from '@ai-sdk/cohere';
 import { rerank } from 'ai';
 
 const { ranking } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: ['doc1', 'doc2', 'doc3', 'doc4', 'doc5'],
   query: 'relevant information',
   topN: 3, // Return only top 3 most relevant documents
@@ -122,7 +122,7 @@ import { cohere } from '@ai-sdk/cohere';
 import { rerank } from 'ai';
 
 const { ranking } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: ['sunny day at the beach', 'rainy afternoon in the city'],
   query: 'talk about rain',
   providerOptions: {
@@ -144,7 +144,7 @@ import { cohere } from '@ai-sdk/cohere';
 import { rerank } from 'ai';
 
 const { ranking } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: ['sunny day at the beach', 'rainy afternoon in the city'],
   query: 'talk about rain',
   maxRetries: 0, // Disable retries
@@ -162,7 +162,7 @@ import { cohere } from '@ai-sdk/cohere';
 import { rerank } from 'ai';
 
 const { ranking } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: ['sunny day at the beach', 'rainy afternoon in the city'],
   query: 'talk about rain',
   abortSignal: AbortSignal.timeout(5000), // Abort after 5 seconds
@@ -179,7 +179,7 @@ import { cohere } from '@ai-sdk/cohere';
 import { rerank } from 'ai';
 
 const { ranking } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: ['sunny day at the beach', 'rainy afternoon in the city'],
   query: 'talk about rain',
   headers: { 'X-Custom-Header': 'custom-value' },
@@ -195,7 +195,7 @@ import { cohere } from '@ai-sdk/cohere';
 import { rerank } from 'ai';
 
 const { ranking, response } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: ['sunny day at the beach', 'rainy afternoon in the city'],
   query: 'talk about rain',
 });
@@ -209,6 +209,8 @@ Several providers offer reranking models:
 
 | Provider                                                                      | Model                                 |
 | ----------------------------------------------------------------------------- | ------------------------------------- |
+| [Cohere](/providers/ai-sdk-providers/cohere#reranking-models)                 | `rerank-v4.0-pro`                     |
+| [Cohere](/providers/ai-sdk-providers/cohere#reranking-models)                 | `rerank-v4.0-fast`                    |
 | [Cohere](/providers/ai-sdk-providers/cohere#reranking-models)                 | `rerank-v3.5`                         |
 | [Cohere](/providers/ai-sdk-providers/cohere#reranking-models)                 | `rerank-english-v3.0`                 |
 | [Cohere](/providers/ai-sdk-providers/cohere#reranking-models)                 | `rerank-multilingual-v3.0`            |

package/docs/03-ai-sdk-core/32-evaluation.mdx

@@ -93,7 +93,7 @@ const registry = createProviderRegistry({
   triage: customProvider({
     evaluationModels: {
       native: typeSafeAi.evaluationModel('jev-latest'),
-      compact: openai.evaluationModel('gpt-5.6-luna'),
+      compact: openai.evaluationModel('gpt-6-luna'),
     },
     fallbackProvider: typeSafeAi,
   }),

package/docs/03-ai-sdk-core/35-image-generation.mdx

@@ -286,7 +286,7 @@ const { image } = await generateImage({
 
 ## Generating Images with Language Models
 
-Some language models such as Google `gemini-2.5-flash-image` support multi-modal outputs including images.
+Some language models such as Google `gemini-3.1-flash-image-preview` support multi-modal outputs including images.
 With such models, you can access the generated images using the `files` property of the response.
 
 ```ts
@@ -294,7 +294,7 @@ import { google } from '@ai-sdk/google';
 import { generateText } from 'ai';
 
 const result = await generateText({
-  model: google('gemini-2.5-flash-image'),
+  model: google('gemini-3.1-flash-image-preview'),
   prompt: 'Generate an image of a comic cat',
 });
 

package/docs/03-ai-sdk-core/37-speech.mdx

@@ -13,7 +13,7 @@ import { generateSpeech } from 'ai';
 import { openai } from '@ai-sdk/openai';
 
 const audio = await generateSpeech({
-  model: openai.speech('tts-1'),
+  model: openai.speech('gpt-4o-mini-tts'),
   text: 'Hello, world!',
   voice: 'alloy',
 });
@@ -38,7 +38,7 @@ import { generateSpeech } from 'ai';
 import { openai } from '@ai-sdk/openai';
 
 const audio = await generateSpeech({
-  model: openai.speech('tts-1'),
+  model: openai.speech('gpt-4o-mini-tts'),
   text: 'Hello, world!',
   providerOptions: {
     openai: {
@@ -59,7 +59,7 @@ import { openai } from '@ai-sdk/openai';
 import { generateSpeech } from 'ai';
 
 const audio = await generateSpeech({
-  model: openai.speech('tts-1'),
+  model: openai.speech('gpt-4o-mini-tts'),
   text: 'Hello, world!',
   abortSignal: AbortSignal.timeout(1000), // Abort after 1 second
 });
@@ -75,7 +75,7 @@ import { openai } from '@ai-sdk/openai';
 import { generateSpeech } from 'ai';
 
 const audio = await generateSpeech({
-  model: openai.speech('tts-1'),
+  model: openai.speech('gpt-4o-mini-tts'),
   text: 'Hello, world!',
   headers: { 'X-Custom-Header': 'custom-value' },
 });
@@ -90,7 +90,7 @@ import { openai } from '@ai-sdk/openai';
 import { generateSpeech } from 'ai';
 
 const audio = await generateSpeech({
-  model: openai.speech('tts-1'),
+  model: openai.speech('gpt-4o-mini-tts'),
   text: 'Hello, world!',
 });
 
@@ -117,7 +117,7 @@ import { openai } from '@ai-sdk/openai';
 
 try {
   await generateSpeech({
-    model: openai.speech('tts-1'),
+    model: openai.speech('gpt-4o-mini-tts'),
     text: 'Hello, world!',
   });
 } catch (error) {

package/docs/03-ai-sdk-core/39-file-uploads.mdx

@@ -26,7 +26,7 @@ const { providerReference } = await uploadFile({
 });
 
 const { text } = await generateText({
-  model: openai.responses('gpt-4o-mini'),
+  model: openai.responses('gpt-6-luna'),
   messages: [
     {
       role: 'user',

package/docs/03-ai-sdk-core/41-skill-uploads.mdx

@@ -174,7 +174,7 @@ Pass the `providerReference` inside the `shell` tool's `environment.skills` arra
 
 ```ts
 await generateText({
-  model: openai.responses('gpt-5.2'),
+  model: openai.responses('gpt-6-astra'),
   tools: {
     shell: openai.tools.shell({
       environment: {

package/docs/03-ai-sdk-core/42-batch.mdx

@@ -142,7 +142,7 @@ const batch = await startBatch({
     {
       id: 'red-panda',
       type: 'image',
-      model: 'gemini-2.5-flash-image',
+      model: 'gemini-3.1-flash-image-preview',
       prompt: 'A red panda reading beside a cabin window',
       aspectRatio: '16:9',
     },

package/docs/03-ai-sdk-core/45-provider-management.mdx

@@ -38,8 +38,8 @@ import {
 export const openai = customProvider({
   languageModels: {
     // replacement model with custom provider options:
-    'gpt-5.1': wrapLanguageModel({
-      model: gateway('openai/gpt-5.1'),
+    'gpt-6-astra': wrapLanguageModel({
+      model: gateway('openai/gpt-6-astra'),
       middleware: defaultSettingsMiddleware({
         settings: {
           providerOptions: {
@@ -51,8 +51,8 @@ export const openai = customProvider({
       }),
     }),
     // alias model with custom provider options:
-    'gpt-5.1-high-reasoning': wrapLanguageModel({
-      model: gateway('openai/gpt-5.1'),
+    'gpt-6-astra-high-reasoning': wrapLanguageModel({
+      model: gateway('openai/gpt-6-astra'),
       middleware: defaultSettingsMiddleware({
         settings: {
           providerOptions: {
@@ -78,8 +78,8 @@ import { customProvider, gateway } from 'ai';
 // custom provider with alias names:
 export const anthropic = customProvider({
   languageModels: {
-    opus: gateway('anthropic/claude-opus-4.1'),
-    sonnet: gateway('anthropic/claude-sonnet-4.5'),
+    opus: gateway('anthropic/claude-opus-5.5'),
+    sonnet: gateway('anthropic/claude-sonnet-5'),
     haiku: gateway('anthropic/claude-haiku-4.5'),
   },
   fallbackProvider: gateway,
@@ -100,10 +100,10 @@ import {
 
 export const myProvider = customProvider({
   languageModels: {
-    'text-medium': gateway('anthropic/claude-3-5-sonnet-20240620'),
-    'text-small': gateway('openai/gpt-5-mini'),
+    'text-medium': gateway('anthropic/claude-sonnet-5'),
+    'text-small': gateway('openai/gpt-6-luna'),
     'reasoning-medium': wrapLanguageModel({
-      model: gateway('openai/gpt-5.1'),
+      model: gateway('openai/gpt-6-astra'),
       middleware: defaultSettingsMiddleware({
         settings: {
           providerOptions: {
@@ -115,7 +115,7 @@ export const myProvider = customProvider({
       }),
     }),
     'reasoning-fast': wrapLanguageModel({
-      model: gateway('openai/gpt-5.1'),
+      model: gateway('openai/gpt-6-astra'),
       middleware: defaultSettingsMiddleware({
         settings: {
           providerOptions: {
@@ -146,7 +146,7 @@ import { customProvider, uploadFile, uploadSkill } from 'ai';
 // custom provider with files interface:
 const myOpenAI = customProvider({
   languageModels: {
-    'gpt-4o-mini': openai.responses('gpt-4o-mini'),
+    'gpt-6-luna': openai.responses('gpt-6-luna'),
   },
   files: openai.files(),
 });
@@ -154,7 +154,7 @@ const myOpenAI = customProvider({
 // custom provider with skills interface:
 const myAnthropic = customProvider({
   languageModels: {
-    sonnet: anthropic('claude-sonnet-4-5'),
+    sonnet: anthropic('claude-sonnet-5'),
   },
   skills: anthropic.skills(),
 });
@@ -224,9 +224,9 @@ import { generateText } from 'ai';
 import { registry } from './registry';
 
 const { text } = await generateText({
-  model: registry.languageModel('openai:gpt-5.1'), // default separator
+  model: registry.languageModel('openai:gpt-6-astra'), // default separator
   // or with custom separator:
-  // model: customSeparatorRegistry.languageModel('openai > gpt-5.1'),
+  // model: customSeparatorRegistry.languageModel('openai > gpt-6-astra'),
   prompt: 'Invent a new holiday and describe its traditions.',
 });
 ```
@@ -295,7 +295,7 @@ import {
 
 const registry = createProviderRegistry({
   openai: customProvider({
-    languageModels: { 'gpt-4o-mini': openai.responses('gpt-4o-mini') },
+    languageModels: { 'gpt-6-luna': openai.responses('gpt-6-luna') },
     files: openai.files(),
   }),
 });
@@ -307,7 +307,7 @@ const { providerReference } = await uploadFile({
 });
 
 const { text } = await generateText({
-  model: registry.languageModel('openai:gpt-4o-mini'),
+  model: registry.languageModel('openai:gpt-6-luna'),
   messages: [
     {
       role: 'user',
@@ -330,7 +330,7 @@ import { createProviderRegistry, customProvider, uploadSkill } from 'ai';
 
 const registry = createProviderRegistry({
   anthropic: customProvider({
-    languageModels: { sonnet: anthropic('claude-sonnet-4-5') },
+    languageModels: { sonnet: anthropic('claude-sonnet-5') },
     skills: anthropic.skills(),
   }),
 });
@@ -356,7 +356,7 @@ Here is an example that implements the following concepts:
 - pre-configure model settings (here: `anthropic > reasoning`)
 - validate the provider-specific options (here: `AnthropicLanguageModelOptions`)
 - use a fallback provider (here: `anthropic > *`)
-- limit a provider to certain models without a fallback (here: `groq > gemma2-9b-it`, `groq > qwen-qwq-32b`)
+- limit a provider to certain models without a fallback (here: `groq > llama-3.1-8b-instant`, `groq > qwen/qwen3.6-27b`)
 - define a custom separator for the provider registry (here: `>`)
 
 ```ts
@@ -419,8 +419,8 @@ export const registry = createProviderRegistry(
     // limit a provider to certain models without a fallback
     groq: customProvider({
       languageModels: {
-        'gemma2-9b-it': groq('gemma2-9b-it'),
-        'qwen-qwq-32b': groq('qwen-qwq-32b'),
+        'llama-3.1-8b-instant': groq('llama-3.1-8b-instant'),
+        'qwen/qwen3.6-27b': groq('qwen/qwen3.6-27b'),
       },
     }),
   },
@@ -462,7 +462,7 @@ globalThis.AI_SDK_DEFAULT_PROVIDER = openai;
 import { streamText } from 'ai';
 
 const result = await streamText({
-  model: 'gpt-5.1', // Uses OpenAI provider without prefix
+  model: 'gpt-6-astra', // Uses OpenAI provider without prefix
   prompt: 'Invent a new holiday and describe its traditions.',
 });
 ```

package/docs/03-ai-sdk-core/60-telemetry.mdx

@@ -235,7 +235,7 @@ import { streamText } from 'ai';
 import { DevToolsTelemetry } from '@ai-sdk/devtools';
 
 const result = streamText({
-  model: openai('gpt-4o'),
+  model: openai('gpt-6-astra'),
   prompt: 'Hello!',
   telemetry: {
     integrations: [DevToolsTelemetry()],

package/docs/03-ai-sdk-core/65-devtools.mdx

@@ -49,7 +49,7 @@ Telemetry is enabled automatically once an integration is registered — no per-
 import { generateText } from 'ai';
 
 const result = await generateText({
-  model: openai('gpt-4o'),
+  model: openai('gpt-6-astra'),
   prompt: 'What cities are in the United States?',
 });
 ```
@@ -61,7 +61,7 @@ import { streamText } from 'ai';
 import { DevToolsTelemetry } from '@ai-sdk/devtools';
 
 const result = streamText({
-  model: openai('gpt-4o'),
+  model: openai('gpt-6-astra'),
   prompt: 'Hello!',
   telemetry: {
     integrations: [DevToolsTelemetry()],
@@ -159,7 +159,7 @@ Telemetry is enabled automatically, but AI SDK 7 excludes raw request and respon
 
 ```ts highlight="4-7"
 const result = await generateText({
-  model: openai('gpt-4o'),
+  model: openai('gpt-6-astra'),
   prompt: 'Hello!',
   include: {
     requestBody: true,

package/docs/03-ai-sdk-core/65-lifecycle-callbacks.mdx

@@ -257,7 +257,7 @@ const { embeddings } = await embedMany({
 });
 
 const { ranking } = await rerank({
-  model: cohere.reranking('rerank-v3.5'),
+  model: cohere.reranking('rerank-v4.0-pro'),
   documents: values,
   query: 'talk about rain',