Anthropic has introduced a beta feature called message threads for its Claude API, designed to improve response times for large context requests. This new parameter, available by default for Claude Code users, reduces the time to first token by approximately 35% for requests between 500,000 and 1 million tokens. While token usage and pricing remain unchanged, threads are stored in memory and expire after a period of inactivity, requiring a full conversation resend if left idle. AI
IMPACT Potentially speeds up interactions with large context windows, benefiting agent-based applications.
RANK_REASON Product update for an existing AI model API.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →