Mistral Large 4 (Le Chonk, open weights)
Mistral Large 4 (Le Chonk, open weights).
15
Benchmarks with a score
0
Ranked first, out of five or more models
0
In the top three
Mistral AI Oct 2026 Mistral Large 4 (Le Chonk, open weights)
Open weights Flagship
No Epoch Capabilities Index score yet
- Legal Research Bench 35.6% #45 /88
- Vals Index v2.2 56.5% #17 /32
- Tax Agent Bench 30.7% #28 /79
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.
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Vibe Code Bench | 81.6% | #26 of 115 | Official leaderboard | |
| FrontierCode 2.0 | 19.7% | #1 of 1 | Model card |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BrowseComp | 84.5% | #18 of 77 | Model card | |
| Terminal-Bench 4.0 | 41% | #13 of 23 | Model card |
Professional work
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Harvey LAB v1.1 (HLAB-Vals-1.1) | 15.8% | #8 of 33 | Official leaderboard | |
| Tax Agent Bench | 30.7% | #28 of 79 | Official leaderboard | |
| Vals Finance Agent v2.1 | 61.7% | #15 of 35 | Official leaderboard | |
| Legal Research Bench | 35.6% | #45 of 88 | Official leaderboard | |
| Vals Finance Index | 57% | #18 of 35 | Official leaderboard | |
| Vals Index v2.2 | 56.5% | #17 of 32 | Official leaderboard | |
| Excel Modeling Benchmark v1.1 | 68.7% | #23 of 37 | Official leaderboard | |
| Vals Legal Index | 67.8% | #23 of 33 | Official leaderboard | |
| Finch (FinWorkBench) | 73.3% | #1 of 1 | Model card |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Artificial Analysis Cyber Index | 51% | #1 of 1 | Model card | |
| Lakera B3 (Backbone Breaker Benchmark) | 97.1% | #1 of 1 | Model card |
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.