Caching & Batch Savings Calculator
Estimate how much prompt caching or the Batch API could save on repeated or high-volume requests, using each provider's published discount rates.
Discount rates last verified 2026-08-10.
Monthly cost comparison
Sources
- OpenAI pricing: https://platform.openai.com/docs/pricing
- Anthropic pricing: https://docs.claude.com/en/docs/about-claude/pricing
- Google pricing: https://ai.google.dev/gemini-api/docs/pricing
Frequently asked questions
When does prompt caching actually help?
When the same large block of text — a system prompt, document, or codebase — is reused across many requests. Caching a prompt you only send once provides no benefit and, for Anthropic, costs slightly more than not caching at all due to the cache-write premium.
When does the Batch API make sense?
For large volumes of requests that don't need an instant response — bulk classification, offline summarization, data processing — where waiting for results (typically up to 24 hours) is acceptable in exchange for a flat discount.
Can I combine caching and batch discounts?
Provider documentation doesn't consistently confirm both discounts stack on the same request — treat this calculator's two scenarios as separate options to compare, not necessarily additive.