Search the web and fetch web pages with Anthropic.

Use OpenAI's built-in file search tool.

Use OpenAI's built-in image generation tool.

Use OpenAI's built-in code interpreter tool.

Use OpenAI's built-in web search tool.

Learn how to generate and edit images with Gemini 2.5 Flash Image.

Learn how to integrate AI SDK with your preferred API server - examples for Express, Fastify, Hono, and more.

In AI SDK 5, the OpenAI provider defaults to the Responses API. The Completions API remains available and fully supported.

Chat SDK is now powered by AI SDK 5, Vercel AI Gateway, and AI Elements.

AI Elements is now open-source.

useChat onFinish now includes isAbort, isDisconnect, and isError flags.

Customize warning handling with the new global logger - disable warnings completely or provide your own handler function.

Override the default file download function with custom logic for authentication, retries, and caching.

Learn how to build a chatbot with resumable streams. ai-sdk.dev/docs/ai-sdk...

Generate images with Gemini 2.5 Flash Image.

Access reasoning text with generateObject.

We want to make AI SDK 5 migration smoother. If you upgraded, what issues did you hit? If you haven’t, what’s holding you back?

Generate speech with ElevenLabs.

Learn how to build a market research agent with Gemini and the AI SDK.

Validate UI messages against your custom schemas for tools, data parts, and metadata.

Access the search queries that OpenAI's web search tool uses via the Responses API.

Stream preliminary tool results

Thank you to everyone who contributed to AI SDK 5.

AI SDK 5 Introducing type-safe chat, agentic loop controls, data parts, speech generation and transcription, Zod 4 support, global provider, and raw request access.

Learn how to use message metadata with useChat to track timestamps, token usage, and more.

Save partial results and handle cleanup properly when a UIMessageStream is aborted.

Automatically migrate to AI SDK 5 with our initial set of codemods.

Generate images with Google's Imagen 3.

You can now modify messages for each step. For multi-step generations, this provides full control over your context.

Use the readUIMessageStream helper for custom stream processing.

Build agents with the AI SDK.

Chat SDK is now using AI SDK 5 Beta.

Our Next.js, Nuxt, and Svelte quickstart guides have been updated to AI SDK 5 Beta. Get started today.

The first set of bug fixes & improvements are now live for AI SDK 5 Beta. We'd love to hear your feedback.

AI SDK 5 is now in Beta! We redesigned the AI SDK to create a next-generation AI framework: • Create type-safe, flexible chats with React, Svelte, Vue • Build agents with prepareStep, stopWhen, and provider-executed tools • Rely on an updated, robust language model spec

Citations are now supported with Cohere models.

Magistral is now available on the AI Gateway. Try it out on the playground.

AI SDK 5.0.0-alpha.12 You can now access raw chunks to use bleeding-edge provider features before they are officially supported.

PDF citations are now available with Anthropic models.

AI SDK 5.0.0-alpha.11 Tool callbacks expose args throughout the generation lifecycle with start, streaming, and completion events.

Use Anthropic's web search tool.

Morph V2, a 1600 tokens/second code-modification model, is now available on the playground.

AI SDK 5 introduces a new language model specification. Provider and middleware maintainers: Your packages will need to be updated for LanguageModelV2 compatibility. The complete specification is now available to help guide your migration.

Use Zod 4 with the AI SDK.

The latest version of DeepSeek R1 (0528) is now available on the Playground.

Use Claude 4 through the AI Gateway.

Use Claude 4 with the AI SDK.

Claude 4 is now available on the Playground.

Use the new v0 model with the AI SDK.

AI SDK 5.0.0-alpha.3 Use continueUntil with streamText.