Anthropic added beta message threads to its Messages API on November 2, 2026. With the `message-threads-2026-08-12` beta header, developers can keep a conversation on Anthropic’s servers and send only new messages in later requests, which the company says can lower time to first token.
Threads expire five minutes after the last request and are deleted no more than one hour after last use. The feature is available through the Claude API, Claude Platform on AWS, Anthropic-managed Claude Models on Google Cloud, and Claude in Microsoft Foundry.
Why it matters
For developers building long-running conversations or agent loops, threads can reduce the messages they must resend; how much this lowers response time in practice is not specified.
