TechnologyCurrent

Prompt caching (Claude API)

Reuse a processed prompt prefix (tools, system, messages) to cut cost and latency on repeated context.

Details

Enable with cache_control, either one top-level field (automatic breakpoints) or on specific blocks (up to 4). Default lifetime is 5 minutes, refreshed on each use; a 1-hour TTL is available at a higher write price.

About this entry

Docs: Documentation ↗
Author: Anthropic
Original source: platform.claude.com/docs/en/build-with-claude/prompt-caching ↗
Added by: Promptu (curated)
Last verified: Nov 5, 2026