Predict the break-even read count for Anthropic prompt caching given the 1.25x / 0.1x rates

Attempt it

Anthropic charges 1.25x the normal input rate on the cache WRITE (the first request that populates a cached prefix) and 0.1x the normal input rate on each cache READ (every subsequent hit on the same prefix). Compute the minimum number of total cache reads N (not counting the write) after which the cumulative cost is strictly LOWER than running the same N+1 requests against an uncached endpoint at 1.0x. Show the inequality and the integer answer.

Quick reference

TL;DR

Cached cost (write + N reads) = 1.25 + 0.1N. Uncached cost (N+1 calls) = N+1. Set cached < uncached: 0.25 < 0.9N, so N > 0.28. Caching wins from the FIRST read (1.35 < 2.0).

Memory aid

Sign in for the mnemonic that makes this stick. See Pro pricing.

Easy to grasp

Picture a coffee shop that charges $1.25 to engrave your favorite order on a card the first time you visit and $0.10 each time you flash that card later. Without the card, every drink is $1.00. After your first refill ($1.25 + $0.10 = $1.35 total) you've spent more than the $1 for a single drink but less than the $2 you'd have spent on two regular drinks. From your second card flash onward, you're saving by a wide margin. The card pays off remarkably fast.

Key concepts

Prompt caching Anthropic API API pricing math

Quick scan · 2–4 min

Everything important, quickly.

3 min: the inequality and its solution + why N=1 is the break-even + what the '2 reads' figure actually measures + TTL implications + Anthropic vs OpenAI cache mechanics + production placement of cache_control.

Real-world examples

Real products, models, and research that use this idea.

Cursor and Anysphere have publicly reported 70-90% input cost reductions on their Claude usage after enabling prompt caching on system prompts and codebase context.
Anthropic's own Computer Use sessions cache the lengthy system prompt across all turns, making the per-turn input cost manageable despite the verbose tool definitions.
RAG-heavy products (Perplexity-style) cache retrieved chunks when the same retrieval result serves multiple users' follow-up turns within the TTL window.
Agent frameworks like LangChain and the OpenAI Agents SDK in 2026 expose cache-control configuration as a first-class parameter on agent system prompts.
Claude Sonnet 4.5 in production deployments commonly runs with cached system prompts + variable user messages; the cached portion bills at 0.1x and dominates the cost savings.

Interviewer follow-ups

What an interviewer would ask next. Try answering before peeking at the approach.

QHow does Anthropic's extended 1-hour TTL change the break-even math?

Sign in or go Pro to see the approach for this follow-up.

QWhy does Anthropic price the write at 1.25x rather than at 1.0x like OpenAI's auto-cache?

Sign in or go Pro to see the approach for this follow-up.

QHow would you instrument an agent loop to track actual caching efficiency?

Sign in or go Pro to see the approach for this follow-up.

60-second cheat sheet

The night-before-the-interview bullets. Scan these on the way to the call.

Sign in for 5 unlocks/day or go Pro for unlimited revision sheets. See Pro pricing.

References

Primary sources. Skim if you want the original framing.