Changelog UpdateMar 24, 2026OpenAI/OpenAI APIAPIFollowTool search in the Responses API, which lets models defer large tool surfaces until runtime to reduce token usage, preserve cache performance, and improve latency.