Estimate what a Claude API workload will cost. Pick a current model, enter how many input and output tokens you expect, and optionally account for prompt caching (up to 90% off cached input) and the Batch API (50% off). All maths runs in your browser — nothing is sent anywhere.
Current Claude API prices
| Model | Input /Mtok | Output /Mtok | Context | Max output |
|---|---|---|---|---|
| Claude Fable 5 | $10 | $50 | 1,000,000 | 128,000 |
| Claude Opus 4.8 | $5 | $25 | 1,000,000 | 128,000 |
| Claude Sonnet 5 | $2* (→ $3) | $10* (→ $15) | 1,000,000 | 128,000 |
| Claude Haiku 4.5 | $1 | $5 | 200,000 | 64,000 |
How the estimate works
- Base cost = (input tokens ÷ 1M × input price) + (output tokens ÷ 1M × output price).
- Prompt caching: the cached share of input is billed at ~10% of the input price (up to 90% off cached input).
- Batch API: applies a flat 50% discount to both input and output for asynchronous jobs.
- Estimates are pre-tax and exclude any plan credits or negotiated rates.
For plan (subscription) pricing rather than per-token API costs, see our Claude pricing guide. For model selection, see the current Claude models reference and the Claude API guide. Official pricing: platform.claude.com/pricing.
Prices verified: 2026-07-13. Independent guide — not affiliated with Anthropic.