rotative
the decoder · November 3, 2026

Anthropic slashes token time by 35% with Message Threads

anthropic's message threads reduces time to first token by 35 percent for long conversations, benefiting coding agents and workflows. this improves efficiency for large requests.

read in 简体中文 · 繁體中文

anthropic has launched message threads as a public beta on its claude platform, significantly improving the efficiency of long conversations. by only transmitting changes since the last message, message threads cuts time to first token by about 35 percent for requests between 500,000 and one million tokens. this feature is particularly beneficial for coding agents, computer-use agents, and workflows that process large volumes of documents or images. the thread remains encrypted in memory between requests and automatically expires after five minutes of inactivity. importantly, token usage remains unchanged, as claude still reads the full conversation history. message threads is available on the claude platform and can also be accessed through aws, microsoft foundry, and google cloud.

read at source ↗

← rotative