Mistral AIMistral Large 3
Best configuration ranks #358 of 610 on the BenchLeader Index at 47.0 ±5.6. Last measured 8 Sept 2026.
- Blended price
- $0.750/M
- $0.500 in · $1.50 out
- Output speed
- 77 tok/s
- First answer
- 1.04 s
- first token 1.04 s
- Context
- 262k
- Overall index47
- Reasoning54
- Coding38
- Agents & tools47
- Knowledge45
- Instruction following39
- Human preference60
- Multimodal48
- Long context43
- Composite41
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 | Trend |
|---|---|---|---|
| LMArena Hard Prompts | 1431#118 | LMArena | |
| GPQA Diamond (AA)not in index | 68.0%#303 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.2%#449 | Artificial Analysis | |
| Kagi LLM Benchmark | 50.9%#83 | Kagi LLM Benchmark |
Coding
| Benchmark | default | Source | Trend |
|---|---|---|---|
| LMArena Coding | 1469#102 | LMArena | |
| LMArena WebDev | 1230#111 | LMArena | |
| SciCode (AA)not in index | 36.6%#130 | Artificial Analysis |
Agents & tools
| Benchmark | default | Source | Trend |
|---|---|---|---|
| Terminal-Bench Hard | 15.9%#184 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 24.6%#289 | Artificial Analysis |
Knowledge
| Benchmark | default | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -39.6#252 | Artificial Analysis |
Instruction following
| Benchmark | default | Source | Trend |
|---|---|---|---|
| IFBench | 36.2%#289 | Artificial Analysis |
Human preference
| Benchmark | default | Source | Trend |
|---|---|---|---|
| LMArena Text | 1414#115 | LMArena |
Multimodal
| Benchmark | default | Source | Trend |
|---|---|---|---|
| LMArena Vision | 1227#64 | LMArena | |
| MMMU-Pro | 55.7%#191 | Artificial Analysis |
Long context
| Benchmark | default | Source | Trend |
|---|---|---|---|
| AA-LCR | 36.0%#302 | Artificial Analysis |
Composite
| Benchmark | default | Source | Trend |
|---|---|---|---|
| AA Intelligence Index | 9.7#321 | Artificial Analysis |
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 | $0.0007 | – | 4.9 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0069 | – | 8.8 s |
| Code edit | 6,000 / 1,500 | $0.0053 | – | 20.6 s |
| Agentic coding session | 60,000 / 4,000 | $0.036 | – | 53.1 s |
| Structured extraction | 2,000 / 200 | $0.0013 | – | 3.6 s |
See also
Data as of 9 Sept 2026. Compare with another model.