AI cost calculator
Claude API pricing calculator
Enter your requests and token counts to see the monthly Claude API bill on each current model at Anthropic's list price. Sonnet 5.5 is $2 per million input tokens and $10 per million output tokens; cached input costs a tenth of that.
Your inputs
Result
$2,130 a month on Claude Sonnet 5.5
$7.10 per 1,000 requests.
- Cost per request
- $0.0071
- Cost per 1,000 requests
- $7.10
- Per month
- $2,130
- Per year
- $25,560
- Saved by prompt caching
- $570.00 a monthagainst the same requests with no cache hits
- Cheapest Anthropic model for this workload
- Claude Haiku 5.5$108.00 a month
Worked examples
Each row is the calculator run on the inputs shown, with every other input at its default.
| Workload | Model | Requests a day | Tokens in / out | Per month |
|---|---|---|---|---|
| Support chatbot | Claude Fable 5.1 | 5,000 | 3,000 / 300 | $751.50 |
| Support chatbot on Haiku 5.5 | Claude Haiku 5.5 | 5,000 | 3,000 / 300 | $39.15 |
| Nightly document batch | Claude Fable 5.1 | 2,000 | 30,000 / 1,500 | $2,250 |
| Coding agent on Opus 5.5 | Claude Opus 5.5 | 2,000 | 60,000 / 2,000 | $4,488 |
| Research agent on Fable 5.1 | Claude Fable 5.1 | 300 | 120,000 / 6,000 | $5,076 |
| Long prompts on Haiku 5.5 | Claude Haiku 5.5 | 1,000 | 150,000 / 800 | $1,298 |
How it is calculated
Each request is billed for its input and output tokens at the model's price per million tokens. Input read from the prompt cache costs 0.1 times the input price (0.025 times on Fable 5.1).
Haiku 5.5 has two prices: $0.10 / $0.50 per million tokens for prompts up to 100,000 tokens and $0.50 / $2.50 above that. The calculator switches to the higher price when your input per request is over 100,000 tokens.
Writing to the cache costs more than plain input: 1.25 times the input price for a 5-minute cache and 2 times for a 1-hour cache. For a prompt that is reused many times a day that one-time cost is small, so it is left out here. The Claude Code calculator counts cache writes because coding sessions write a lot of new context.
Claude 4.7 and later models use a tokenizer that makes about 30% more tokens for the same text. If your token counts come from an older Claude model, raise them by about 30% or use the tokenizer change calculator.
Questions
How much does the Claude API cost?
Per million tokens: Haiku 5.5 is $0.10 input and $0.50 output, Sonnet 5.5 is $2 and $10, Opus 5.5 is $4 and $20, and Fable 5.1 is $10 and $50. Cached input is 0.1 times the input price (0.025 times on Fable 5.1).
How much does the Batch API save?
The Batch API is 50% off both input and output tokens. Requests are processed asynchronously, usually within 24 hours, which suits nightly jobs, evaluations and bulk document work.
Is the Claude API cheaper than a Claude subscription?
It depends on volume. The API is pay per token with no monthly fee; Pro and Max are flat monthly prices with usage limits. The Claude Pro vs Max calculator compares the two for your usage.
What does US-only inference cost?
Setting inference_geo to us keeps inference in the United States and multiplies token prices by 1.1.
Which Claude model is cheapest?
Haiku 5.5, at $0.10 per million input tokens for prompts up to 100,000 tokens. The chart above ranks every Claude model for your exact workload.
Sources
- Anthropic: Claude API pricing, checked 2026-10-29
Prices and plans change. This page was last checked on 2026-10-29. The data is open: llm-prices.json (CC BY 4.0).