DigestAI news desk

Cut through the AI noise.

Agents & Tools3 min read

Anthropic adds message threads to Claude Platform for faster agent responses

Anthropic launched message threads in public beta on November 2, 2026, for its Claude Platform and Claude Models on AWS, Microsoft Foundry, and Google Cloud. The feature lets developers send only new messages in long-running agent sessions—such as coding or document-heavy workflows—reducing median latency for large requests (500,000 to 1 million tokens) by about 35% by avoiding full-conversation…

1 source

Key points

  • Message threads cut latency by 35% for large requests (500K–1M tokens) by skipping full-conversation re-uploads
  • Beta available on Claude Platform, AWS, Microsoft Foundry, and Google Cloud with all models enabled
  • Threads expire in 5 minutes, require manual conversation history backup, and exclude HIPAA/zero-retention use

Threads work by splitting requests into create (full conversation) and continue (new messages only) steps. They expire after five minutes of inactivity and are stored in memory under server encryption, deleted within one hour. Developers must handle their own conversation history since there’s no API to retrieve or delete threads. The beta excludes organizations with Inference hooks, Message Batches, or zero-data-retention requirements. Anthropic’s SDK simplifies thread management but does not auto-retry failed continues or cache prompts—developers must manually implement caching and error handling.

The story so far

3 episodes →
  1. Anthropic adds message threads to Claude Platform for faster agent responsesthis story
Full story from Unite.AI · by Jonas Reeve, Cognitive AI & AGI, AI Research AgentOpen source ↗

Anthropic Launches Message Threads Beta for Long-Running Agents

Unite.AI · 2 November 2026

Loading the full article…

This text was published by Unite.AI and written by Jonas Reeve, Cognitive AI & AGI, AI Research Agent. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗

Topics · follow one to build your own front page

The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.

Comments

via GitHub Discussions

More in Agents & Tools

All →

Related stories