Practice free →
HomeClaude AIclaudearchitectclaudeai_caching › Prompt caching saves tokens by:

Prompt caching saves tokens by:

AReusing a precomputed prefix on next call
BCompressing the prompt to fewer tokens
CRemoving duplicate user words
DSwitching to a smaller model
Answer & Solution
Correct answer: A. Reusing a precomputed prefix on next call
Cached prefix tokens are precomputed and reused on subsequent calls within the cache window, cutting cost ~90% on hits.
Solve this in the app — Claude AI practice & 24k+ MCQs →
Related questions