Anthropic's Message Threads cut time to first token by 35 percent for long conversations
decoder ·
Anthropic has launched Message Threads as a public beta feature on its Claude platform to optimize long conversations. Previously, every API request required resending the entire conversation history, which was inefficient for lengthy interactions. The new system only transmits changes since the last message, resulting in a 35 percent reduction in time to first token for requests between 500,000 and one million tokens. This improvement particularly benefits coding agents, computer-use agents, and workflows that process large volumes of documents or images. Although transmission efficiency improves, total token usage remains unchanged because Claude still reads the full conversation history internally. To ensure security, threads are encrypted in memory between requests and automatically expire after five minutes of inactivity. The feature is accessible via the Claude platform and through major cloud providers including AWS, Microsoft Foundry, and Google Cloud.
decoder