Affected version
Reports that share this version, grouped in one feed.
When using astream_events version='v3' with create_agent, the final response from the top-level agent is not streamed token-by-token; all tokens arrive at once with identical timestamps, while intermediate outputs stream correctly. This impacts real-time streaming applications.
The OpenAI Responses API excludes deferred tool schemas from billed input only when they are inside a `namespace` block. `ProviderToolSearchMiddleware` emits flat `function` tools with `defer_loading: true` plus `tool_search`, causing OpenAI to bill the full schemas and adding overhead, so token usage increases instead of decreasing.
LangChain, LlamaIndex, and Pydantic
Pinecone, Chroma, Weaviate, Qdrant, and neighboring stores
Next.js, Supabase, and the surrounding app stack