aiqbench analysis
This article details a practical optimization for using Anthropic's Claude Code by adjusting cache lifetime settings to capture significant cost reductions following recent pricing changes. By extending the cache window beyond the default five minutes, developers can reduce token expenses by 10–15% by minimizing the frequency of expensive full-history re-reads during agentic coding tasks.
Claude Code Prompt Caching: Cut Token Costs 10–15%

From Larridin
ExcerptWhen Anthropic recently cut the price of a Claude Fable 5.1 cache read by 75%, our CTO, Ameya Kanitkar, noticed we weren't collecting most of that discount. One small config change later, our Claude Code token costs dropped 10 to 15% overnight.
Ameya Kanitkar, co-founder and CTO of Larridin, on LinkedIn: Fable 5.1 cut the price of a cache read by 75%. We changed one setting in Claude Code to actually collect the savings, and our token costs dropped 10 to 15% overnight. The setting is a cache lifetime. To see why one line of config moves the bill that much, look at where the money goes. The biggest line on a Claude Code bill is the model re-reading what it already read.
Excerpt shown under the publisher's RSS terms · 123 words · full text and figures are on the publisher's site
Related coverage

How Much Does Claude Code Cost? Real Session Costs by Model
OpenRouter · Oct 30, 2026

Incident Report: Prompt Cache Invalidation for Claude Code on Bedrock Invoke
LiteLLM Blog · Jul 13, 2026

Optimizing cost and latency with Amazon Bedrock prompt caching
AWS · Sep 15, 2026
Anthropic releases Claude Haiku 5.5 small model and halves Sonnet 5.5 cache read prices
SiliconANGLE · Oct 7, 2026