Anthropic 推出 Message Threads 公开测试版，长对话首 token 延迟缩短约 35%
Anthropic's Message Threads cut time to first token by 35 percent for long conversations
- 1Anthropic 在 Claude 平台以公开测试版推出 Message Threads，API 请求只传增量而非整段对话历史
- 2官方称 50 万至 100 万 token 请求的首 token 时间缩短约 35%（官方口径，暂无独立验证）
- 3线程在请求间加密保存于内存，闲置 5 分钟后过期，token 用量不变，Claude 仍读取完整历史
- 4功能在 Claude 平台及 AWS、Microsoft Foundry、Google Cloud 可用，编码与计算机操作类智能体受益最大
Message Threads 把长对话状态改为服务端保存、只传增量，实质是把智能体长会话的延迟问题交给基础设施层解决。
- 影响：编码、计算机操作和文档密集型工作流的响应速度有望改善；token 计费不变，收益体现在速度而非账单成本；线程闲置 5 分钟过期，长会话应用需要处理线程重建逻辑。
- 参考：可考虑先统计自己业务对话的平均 token 规模，判断是否落在官方收益区间；值得在编码或文档处理场景做小规模延迟对比，并检查现有链路对线程过期后的重建处理。
- 依据：35% 的首 token 时间降幅为 Anthropic 官方口径，暂无独立验证；5 分钟过期与 token 用量不变等细节出自官方文档。
Anthropic released Message Threads as a public beta on its Claude platform. Until now, every API request had to resend the entire conversation history. Message Threads only transmits what's changed since the last message, cutting time to first token by about 35 percent for requests between 500,000 and one million tokens, according to Anthropic.
Coding agents, computer-use agents, and workflows processing lots of documents or images benefit the most, the company says. The thread stays encrypted in memory between requests and expires after five minutes of inactivity. Token usage doesn't change since Claude still reads the full conversation history. The feature works on the Claude platform and through AWS, Microsoft Foundry, and Google Cloud. Details are in the documentation.
AI News Without the Hype – Curated by Humans
Subscribe to THE DECODER for ad-free reading, a weekly AI newsletter, our exclusive "AI Radar" frontier report six times a year, full archive access, and access to our comment section.
来源：The Decoder：AI News · the-decoder.com