Anthropic adds message threads to the Claude API so agents can stop resending entire conversations

On Monday, Anthropic launched message threads in public beta for developers building agents on the Claude API. Today, each request an agent makes normally carries every earlier message and tool result. A coding agent can make hundreds of requests in one session, so once that history reaches hundreds of thousands of tokens, resending it adds a delay before Claude starts to reply. With threads, an agent sends the full conversation once. After that, it sends only new messages plus the ID of Claude's last reply. Anthropic has already turned threads on for most users of Claude Code, its coding agent. On requests of 500,000 to 1 million tokens, the company says the typical wait before Claude starts answering fell by about 35%. The feature saves time but not money: Claude still reads the whole conversation on every turn, and prices and rate limits stay the same. For costs, Anthropic points to the 75% cut in cache prices it made in September. Threads are held in memory, encrypted and never written to disk. They expire five minutes after the last request, so developers still need to keep their own copy of the conversation. The feature targets long, tool-heavy agent sessions, where repeated uploads slow responses.