Mistral AIAuto-detectedMistral Large
Mistral Large is a Mistral AI proprietary model, released 26 Feb 2024. Its best configuration ranks #514 of 372 on the BenchLeader Index at 40.8 ±4.6, in the lower half. It scores highest in coding (41) and lowest in maths (30). At $6.00 per million tokens blended it is among the most expensive ranked models. Last measured 2 Sept 2026.
- Blended price
- $6.00/M
- $4.00 in · $12.00 out
- Output speed
- –
- First answer
- –
- Context
- 32k
- Overall index41
- Reasoning34
- Coding41
- Maths30
- Human preference39
Versions
Mistral AI has shipped 3 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Mistral Largethis page | 26 Feb 2024 | 40.8 | #514 |
| Mistral Large 3 | – | 46.8 | #368 |
| Mistral Large 2 | – | 39.5 | #540 |
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source |
|---|---|---|
| GPQA Diamond | 38.8%#241 | Epoch AI Benchmarking Hub |
| LMArena Hard Prompts | 1262#270 | LMArena |
Coding
| Benchmark | default | Source |
|---|---|---|
| LMArena Coding | 1295#265 | LMArena |
Maths
| Benchmark | default | Source |
|---|---|---|
| OTIS Mock AIME | 1.9%#247 | Epoch AI Benchmarking Hub |
| MATH Level 5 | 24.5%#72 | Epoch AI Benchmarking Hub |
Human preference
| Benchmark | default | Source |
|---|---|---|
| LMArena Text | 1242#276 | LMArena |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0052 | – | – |
| Summarise a 30-page report | 12,000 / 600 | $0.055 | – | – |
| Code edit | 6,000 / 1,500 | $0.042 | – | – |
| Agentic coding session | 60,000 / 4,000 | $0.288 | – | – |
| Structured extraction | 2,000 / 200 | $0.010 | – | – |
See also
Data as of 10 Sept 2026. Compare with another model.