The importance of adapting quickly in an ever-changing AI world.
We’ve been listening to your feedback and working hard to expand the capabilities of the AI SDK while improving its existing functionality. Today, we’re launching AI SDK 3.2.
This release advances the SDK across four major areas:
Agents : Extended generateText and streamText for multi-step workflows
Providers : Added new providers and expanded capabilities for Anthropic and Google models
Embeddings : Introduced embeddings support to power use cases like retrieval augmented generation (RAG) and semantic search
DX improvements : Improved AI SDK observability and enabled client-side tool calls
Link to heading Agents
As AI models become more capable, they’ve been charged with increasingly complex tasks — some of which require multiple steps.…