[ ABORT TO HUD ]
SEQ. 1
SEQ. 2

Thresholds, Costs, ZDR

⚡ Prompt Caching Framework⏱ 15 min⭐ 400 BASE XP⌨ HANDS-ON LAB

Economic and Technical Limits

Caching involves a write premium (1.25x base input cost) during the initial serialization of the context. However, all subsequent 'reads' of that cache hit a massive 90% discount (0.1x base input cost). To justify the overhead, Anthropic enforces minimum token thresholds:

  • 1,024 tokens: Claude 3.7 Sonnet, Claude 3.5 Sonnet, Claude 3 Opus
  • 2,048 tokens: Claude 3.5 Haiku

If your context prefix is smaller than this minimum threshold, the cache header is ignored.

Lifecycle & Compliance

Caches are ephemeral and have a default TTL (Time To Live) of 5 minutes. Every time a cache is 'read', the TTL timer resets. For enterprises focused on security, this system is fully compatible with Zero Data Retention (ZDR)-the cached bits are held recursively in the inference boundary and vaporize immediately upon expiry, ensuring no persistent logs are generated natively.

⌨ HANDS-ON LABVerify Cache Hit Metrics & Discounts
⭐ +200 XP

Inspect cache read metrics. Fire a follow-up request matching the cached prefix and verify cache_read_input_tokens in usage.

1Re-send the request with the identical cached system prefix and extract the usage metrics to confirm the 90% cache read discount.
lab-sandbox — simulated environment
INFINITY LAB SANDBOX v2.6 — simulated shell
Type the command for the current objective. Helpers: "hint", "solution", "clear".
$
OBJECTIVE 1 / 1 — type "hint" if stuck
SYNAPSE VERIFICATION
QUERY 1 // 3
If you send a 500-token prompt with a cache header to Claude 3.7 Sonnet, what happens?
It caches successfully
The system ignores the header because the context is below the 1,024-token minimum
It costs double tokens
It throws a 400 error