AI Prompt Caching Cost Comparison
By Paige Gilmore, Founder, NetLift · Published 2026-07-28 · Updated 2026-08-11
Prices checked 2026-08-16 against official vendor pricing pages.
Prompt caching reduces API expenses by charging a lower 'Read' rate for reused context after an initial 'Write' fee. For high-volume tasks, Haiku 4.5 offers the lowest entry point at $0.10 per 1M read tokens.
Prompt caching allows engineering teams to store frequently used context—such as large document libraries or codebase summaries—and pay a reduced rate for subsequent requests. Across the Claude model family, costs are split into higher 'Write' fees to store the data and significantly lower 'Read' fees to retrieve it.
Haiku 4.5 provides the most aggressive savings with a Read price of $0.10 per 1M tokens. For advanced reasoning, Sonnet 5 and Opus 5 offer higher performance tiers at $0.20 and $0.50 per 1M read tokens, respectively. These rates provide a path to scale long-context applications without linear cost increases.
Official per-token prices
| Model / rate | Price | Unit |
|---|---|---|
| Claude API — Haiku 4.5 (Prompt Caching Read) | $0.1 | per 1M input tokens |
| Claude API — Sonnet 5 (Prompt Caching Read) | $0.2 | per 1M input tokens |
| Claude API — Opus 5 (Prompt Caching Read) | $0.5 | per 1M input tokens |
| Claude API — Fable 5 (Prompt Caching Read) | $1 | per 1M input tokens |
| Claude API — Haiku 4.5 (Prompt Caching Write) | $1.25 | per 1M input tokens |
| Claude API — Sonnet 5 (Prompt Caching Write) | $2.50 | per 1M input tokens |
| Claude API — Opus 5 (Prompt Caching Write) | $6.25 | per 1M input tokens |
| Claude API — Fable 5 (Prompt Caching Write) | $12.50 | per 1M input tokens |
How do write and read fees impact margins?
Caching introduces a two-tier cost structure. You pay a premium to commit data to the cache, such as $12.50 per 1M tokens for Fable 5. Once cached, every subsequent request that references that data is charged the 'Read' rate, which for Fable 5 is $1.00 per 1M tokens. This makes caching highly effective for applications with high repeat usage of the same background data.
Which model is most efficient for high-volume caching?
Haiku 4.5 is the most cost-effective option for developers, with a Write price of $1.25 and a Read price of $0.10 per 1M tokens. If your workflow requires the higher reasoning capabilities of Opus 5, the Read cost increases to $0.50 per 1M tokens. Sonnet 5 sits in the middle, offering a balance of performance with a Read price of $0.20 per 1M tokens.
To determine if caching is delivering a true return, you must track the cache hit rate against the total time saved by end-users. NetLift helps you map these infrastructure savings back to business value, ensuring that lower API unit costs translate into a measurable net return on your AI adoption rather than just a technical optimization.
Sources
All pricing on this page comes from official vendor pages:
- https://claude.com/pricing — retrieved 2026-08-11
Frequently asked questions
What is the cheapest Claude model for prompt caching?
Haiku 4.5 is the most affordable, charging $1.25 per 1M tokens to write to the cache and $0.10 per 1M tokens to read from it.
How much does it cost to cache data using Claude Opus 5?
On both the Claude API and Claude Platform, Opus 5 costs $6.25 per 1M tokens for writing to the cache and $0.50 per 1M tokens for reading.
What are the caching rates for Claude Sonnet 5?
Sonnet 5 charges $2.50 per 1M tokens to write data to the cache and $0.20 per 1M tokens to read from the cache.
Is Fable 5 more expensive than Sonnet 5 for caching?
Yes. Fable 5 is significantly more expensive, with a Write price of $12.50 and a Read price of $1.00 per 1M tokens, compared to Sonnet 5's $2.50 Write and $0.20 Read prices.
About the author
Paige Gilmore is the founder of NetLift, the AI Value Management platform that helps organisations measure the cost, savings and return of AI adoption. Paige Gilmore on LinkedIn