Mistral Medium 3.1
Mistral Medium 3.1 places in the top three on 2 benchmarks.
41
Benchmarks with a score
0
Ranked first, out of five or more models
2
In the top three
Mistral AI Aug 2025 Mistral Medium 3.1
Closed weights
No Epoch Capabilities Index score yet
- CorpFin v2 50.7% #96 /119
- LMCA 19.1% #114 /147
- TaxEval v2 70.3% #85 /129
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LMArena (Elo) | 1,408 | #124 of 335 | Official leaderboard | |
| Creative Writing v3 | 66.3 | #65 of 120 | Official leaderboard | |
| Lech Mazur Creative Writing | -0.43 | #29 of 50 | Official leaderboard | |
| StudyArena Text | 1,214 | #59 of 85 | Official leaderboard | |
| SnorkelSpatial | 14.6% | #20 of 27 | Official leaderboard | |
| Longform Creative Writing (EQ-Bench) | 46.2 | #91 of 119 | Official leaderboard | |
| ConStory-Bench | 1.35 | #11 of 14 | Official leaderboard | |
| Vectara Hallucination Leaderboard (HHEM) | 11.4% | #134 of 148 | Official leaderboard | |
| Vectara Hallucination Leaderboard (2025 dataset) | 22.7% | #101 of 104 | Official leaderboard | |
| DesignArena ASCII Art | 992 | #76 of 77 | Official leaderboard | |
| Extended NYT Connections | 6.50% | #102 of 103 | Official leaderboard |
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| DTBench | 67.6% | #121 of 180 | Official leaderboard |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| DesignArena Website | 1,126 | #109 of 151 | Official leaderboard | |
| DesignArena 3D Design | 1,064 | #102 of 136 | Official leaderboard | |
| DesignArena Code | 1,111 | #108 of 143 | Official leaderboard | |
| DesignArena UI Component | 1,091 | #105 of 139 | Official leaderboard | |
| DesignArena Data Visualization | 1,142 | #93 of 118 | Official leaderboard | |
| DesignArena Game Dev | 1,064 | #113 of 143 | Official leaderboard | |
| WeirdML | 33.1% | #96 of 112 | Official leaderboard | |
| DesignArena SVG | 983 | #96 of 103 | Official leaderboard | |
| ALE-Bench | 701 | #123 of 126 | Official leaderboard |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Galileo Agent Leaderboard v2 | 61% | #2 of 21 | Official leaderboard | |
| Step Race | 1 | #40 of 59 | Official leaderboard | |
| Elimination Game | 0 | #56 of 56 | Official leaderboard |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| TaxEval v2 | 70.3% | #85 of 129 | Official leaderboard | |
| CorpFin v2 | 50.7% | #96 of 119 | Official leaderboard | |
| Vals Finance Agent v1 | 8.85% | #41 of 47 | Official leaderboard | |
| LegalBench | 62% | #128 of 138 | Official leaderboard | |
| MortgageTax | 36.5% | #84 of 87 | Official leaderboard |
Vision, audio and robotics
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Vision Arena | 1,160 | #97 of 136 | Official leaderboard | |
| ClockBench | 10.6% | #29 of 37 | Official leaderboard | |
| MeasureBench Real-world | 10.6% | #16 of 19 | Official leaderboard | |
| MeasureBench Synthetic | 8.50% | #16 of 19 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BioTIER Permit | 99.9% | #2 of 69 | Official leaderboard | |
| Lakera B3 (Backbone Breaker Benchmark) | 44.1% | #31 of 35 | Official leaderboard | |
| Spiral-Bench v1.2 | 15.0 | #20 of 21 | Official leaderboard | |
| BioTIER Refuse | 10.3% | #67 of 69 | Official leaderboard | |
| Human Pathogen Capabilities Test v2 (HPCT v2) | 24.3% | #70 of 71 | Official leaderboard |
Other
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| LMCA | 19.1% | #114 of 147 | Official leaderboard | |
| Conceptual Reasoning Index | 28.4 | #119 of 151 | Official leaderboard | |
| ACCoRD | 39.2 | #133 of 152 | Official leaderboard |