Mistral API Pricing
8 models · From $0.150/1M to $16.00/1M input
The cheapest Mistral model to run is Ministral 3 8B at $0.150 input and $0.150 output per 1M tokens. On a RAG answer of 8,000 tokens in and 600 out, Mistral models range from $0.0013 (Ministral 3 8B) to $0.128 (Voxtral TTS) per request. The largest context window is Mistral Large 4 at 1.0M tokens. Count tokens with the Mistral token counter.
| Model | Input /1M | Output /1M | RAG answer |
|---|---|---|---|
| Mistral Small 4 | $0.150 | $0.600 | $0.0016 |
| Ministral 3 8B | $0.150 | $0.150 | $0.0013 |
| Ministral 3 14B | $0.200 | $0.200 | $0.0017 |
| Codestral | $0.300 | $0.900 | $0.0029 |
| Mistral Large 3 | $0.500 | $1.50 | $0.0049 |
| Mistral Large 4 | $0.700 | $2.10 | $0.0069 |
| Mistral Medium 3.5 | $1.50 | $7.50 | $0.017 |
| Voxtral TTS | $16.00 | <$0.0001 | $0.128 |