Anthropic launches message threads in public beta on the Claude Platform
Overview
Anthropic has released message threads in public beta on the Claude Platform, so agents send only new messages each turn instead of the full conversation history.
The Claude Blog says median time to first token fell about 35% for most Claude Code users with the feature on by default, on requests of 500k to 1M tokens. A separate X post from ClaudeDevs says the Messages API caches conversations on the server between turns, and that the median request in long Claude Code conversations dropped from about 800 KB to 1.8 KB.
Threads are held in memory for 5 minutes after their last request. Claude still reads the full conversation each turn, so token usage is unchanged. Developers enable the feature by adding a thread parameter and beta header to requests. It is available on the native Claude Platform, Claude Platform on AWS, Claude in Microsoft Foundry, and Claude Models on Google Cloud (Anthropic managed), at standard Claude token rates and existing rate limits.
Written by AI from the articles below · updated Nov 2, 2:31 PM ET
Check the sources:
Developments
2 developments
- Nov 2, 2:06 PM ET · 1 articleAnthropic adds a thread parameter to Claude API requests, available on multiple platformsClaudeDevs: Anthropic adds a thread parameter for Claude message threads
- Nov 2, 12:00 AM ET · 3 articlesAnthropic launches message threads in public beta on the Claude PlatformClaude Blog: Claude Platform adds message threads to cut latency in long-running agents
Article timeline
The articles in this story. Times are ET.
ClaudeDevs@ClaudeDevsOfficialAnthropic adds a thread parameter for Claude message threadsAIAnthropic says developers can enable message threads by adding a thread parameter and beta header to their requests. The feature is available on the Claude Platform, Claude Platform on AWS, Claude in Microsoft Foundry, and Claude on Google Cloud at standard token rates.
ClaudeDevs@ClaudeDevsOfficialClaude Code's message threads cut first-token latency about 35% on long requestsAIAnthropic says its recently integrated message threads in Claude Code reduce median time to first token by about 35% on requests with 500K to 1M tokens. In long conversations, the median request size dropped from about 800 KB to 1.8 KB.
ClaudeDevs@ClaudeDevsOfficialAnthropic's Messages API adds public beta message threads with server-side cachingAIAnthropic says message threads are now in public beta on the Messages API. The API caches conversations on the server between turns, so developers send only new messages plus the previous message ID instead of the full conversation, and Claude responds sooner.

- Claude BlogOfficialPickClaude Platform adds message threads to cut latency in long-running agents
AIAnthropic has released message threads in public beta on the Claude Platform, letting agents send only new messages each turn instead of the full conversation history. For requests with 500k to 1M tokens, median time to first token dropped about 35% for most Claude Code users with the feature on by default. Threads are held in memory for 5 minutes after their last request, and token usage stays the same because Claude still reads the full conversation each turn.
Heat Trend
Not enough continuous observations to show a trend yet.