Anthropic launched message threads on Monday, a public beta feature for the Claude Messages API that lets AI agents stop resending their full conversation history on every turn. Anthropic says the change cuts median time to first token by about 35% on requests between 500,000 and 1 million tokens, a figure drawn from Claude Code users, who already have the feature switched on by default.
The mechanics are simple. Developers add a thread parameter to their Messages API requests. The first request carries the full conversation and opens a thread; every request after that carries only the new messages and the ID of Claude's previous response.
The feature saves time, not tokens. Anthropic says Claude still reads the full conversation on every turn, so token usage stays the same, and standard token rates and existing rate limits apply. The announcement describes no separate charge for using threads.
Key facts
- 01Anthropic opened message threads as a public beta on the Claude Messages API on November 2, 2026.
- 02After the first request, agents send only new messages plus the ID of Claude's previous response.
- 03Median time to first token fell about 35% on 500k–1M token requests, measured on Claude Code users.
- 04Threads expire 5 minutes after the last request and are held encrypted in memory, not written to disk.
- 05Token prices and rate limits are unchanged; Claude still reads the full conversation every turn.
Threads are short-lived by design. A thread expires five minutes after its last request, after which the client has to resend the full conversation to start a new one. Developers therefore still need to keep their own copy of each conversation rather than treating the thread as storage.
On data handling, Anthropic says thread data is encrypted and held in memory between turns, and is not written to disk. Threads are not eligible for zero data retention or HIPAA configurations, according to reporting on the launch.
Anthropic is pitching message threads at three kinds of workloads: long-running coding agents, computer use agents where screenshots stay attached to the thread, and document- and image-heavy workflows. Those are the sessions where context grows into hundreds of thousands of tokens and resending the whole history on each turn adds the most delay.
Message threads works with all Claude models and is available on the native Claude Platform, Claude Platform on AWS, Claude in Microsoft Foundry, and Claude models on Google Cloud managed by Anthropic. Documentation for the parameter sits in the build-with-Claude section of Anthropic's developer platform.
The launch follows Anthropic's release of Opus 5.5 in September and a 75% cut to cache prices in the same month. Message threads targets a different cost than either move: the wait before an agent with a very long context starts responding, which Anthropic measured at about 35% shorter for its own Claude Code users.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
