Head to head
GPT-5 mini vs Mistral Large 3
On a blended workload of one million input tokens and three million output tokens, Mistral Large 3 is 1.3× cheaper than GPT-5 mini.
| GPT-5 mini | Mistral Large 3 | |
|---|---|---|
| Provider | OpenAI | Mistral |
| Input / MTok | $0.25 | $0.5 |
| Output / MTok | $2 | $1.5 |
| Cache read / MTok | $0.025 | — |
| Context window | — | — |
| Blended 1M in + 3M out | $6.25 | $5.00 |
| Status | ga | ga |