Unverified50% confidenceBenchmarkExact time
The cost of asking questions repeatedly after writing back to the wiki is basically on par with cold-reading the source code (28 seconds vs 33 seconds, $0.59 vs $0.62), because the query command still has to verify references against the source files before answering.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOpenAI's official example shows that one write plus nine full reads totals 2.15× the prefix input price, compared to 10× without caching.65% similarUnverifiedIn the session assumed by the article, a gateway labeled "50% off" but with cache reads at only a 20% discount and cache writes at full price actually gives a 28% discount, costing $12.13.65% similarUnverifiedThe price of prompt cache reads is only 0.05×–0.1× the normal input price (depending on the model), the Batch API enjoys a 50% discount when you can wait, and Anthropic charges an additional $10 per 1,000 searches for web search.65% similarUnverifiedA Prompt Caching cache hit read costs only 0.1 times the base input price, while the initial write cost is 2 times the regular input price.64% similarVerifiedThe cache read mechanism retains the KV cache for recurring long prompt prefixes, charging at an extremely low rate upon a cache hit; the leaked cache read price for Opus 5.5 is approximately 1/20 of the input price.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1032437API
curl https://kongchang.com/api/v1/knowledge/claims/1032437MCP
get_claim(id=1032437)