Le Chonk AI pricing & API cost calculator
Estimate the tokens you actually send and receive.
API rates in this snapshot
| USD / 1M tokens | Standard | Launch promotion |
|---|---|---|
| Uncached input | $1.36 | $0.68 |
| Cached input | $0.14 | $0.07 |
| Output | $4.18 | $2.09 |
Compare Mistral Large 4 model documentation and Artificial Analysis launch evaluation, then verify your account billing in Mistral Studio.
Calculate your usage
Everything calculates locally. No inputs are stored or sent. Request count can represent a day, month or entire job; choose your own period.
How the estimate works
Per request = ((input − cached) × input rate + cached × cache rate
+ output × output rate) / 1,000,000
Total = per-request cost × request countTokens are model text units, not words. Prompts, history and tool results contribute to input. Image and document tokenization depends on preparation and the endpoint. Prefer billed usage to word-count estimates.
Three planning examples
| Standard-rate scenario | Input / cached / output per request | Requests | Total |
|---|---|---|---|
| Short chat | 1,000 / 0 / 250 | 1,000 | $2.405 |
| Document review | 50,000 / 0 / 2,000 | 100 | $7.636 |
| Agent loop | 20,000 / 10,000 / 3,000 | 500 | $13.77 |
Invented planning workloads, not model measurements. Repeating a prompt does not by itself prove a billable cache hit.
Costs outside this calculator
- Taxes, exchange fees, third-party hosting and separate chat subscriptions.
- OCR, search, storage, tools and your application infrastructure.
- Retries and extra agent steps unless included in your request count.
- Cache creation or retention charges outside the three snapshot rates: check account terms.
- Self-hosting equipment, power and engineering. Open weights do not make inference free.
Sources & verification
Research snapshot: . Source statements are dated; API access and prices can change.
Original model runs are marked “not run” unless a trace is provided. Read our editorial method.