Mistral Large 4
Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work
Shogo Agent Index
UnrankedNot ranked yet: 1 agentic benchmark measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.
Agentic coding
48
estimated
Long-horizon autonomy
53
estimated
Knowledge-work agents
39
32% measured
Computer use
32
estimated
Security & context learning
45
estimated
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 75 | |
| Terminal-Bench 2.0 | est. | 8% | predicted 64 | |
| DeepSWE | est. | 6% | predicted 43 | |
| FrontierSWE | est. | 5% | predicted 12 | |
| CursorBench | est. | 4% | predicted 35 | |
| GSO-Bench | est. | 4% | predicted 23 | |
| METR Time Horizons | est. | 10% | predicted 73 | |
| Vending-Bench 2 | est. | 6% | predicted 38 | |
| PostTrainBench | est. | 4% | predicted 23 | |
| APEX-Agents | 51.6% | 8% | source | |
| GDPval | est. | 7% | predicted 43 | |
| Remote Labor Index | est. | 5% | predicted 3 | |
| DeepResearch Bench | est. | 5% | predicted 47 | |
| OSWorld 2.0 | est. | 6% | predicted 7 | |
| OSWorld | est. | 4% | predicted 70 | |
| Cybench | est. | 5% | predicted 69 | |
| CL-bench | est. | 5% | predicted 21 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| Chess Puzzles | Reasoning | 33.0% | source |
| CritPt | Reasoning | 11.1% | source |
| DTBench | Knowledge | 88.2% | source |
| FrontierCode | Coding | 39.4% | chisel |
| FrontierMath Tiers 1-3 | Math | 58.6% | source |
| FrontierMath Tier 4 | Math | 26.8% | source |
| GPQA Diamond | Reasoning | 90.2% | source |
| LMCA | Knowledge | 35.8% | source |
| MirrorCode | Coding | 40.0% | source |
| OTIS Mock AIME | Math | 99.4% | source |
| ProofBench | Math | 10.0% | source |
| SciCode | Coding | 54.2% | source |
| SimpleQA Verified | Knowledge | 22.4% | source |
| WeirdML v3 | Coding | 4.2% | opencode |