Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
general
| Benchmark | Mistral Large 3 | Qwen2.5 VL 72B Instruct |
|---|---|---|
| MMMLU | 74.2 | — |
multimodal
| Benchmark | Mistral Large 3 | Qwen2.5 VL 72B Instruct |
|---|---|---|
| MMMU | — | 70.2 |
| MMMU-Pro | — | 51.1 |
| ScreenSpot Pro | — | 43.6 |
agent
| Benchmark | Mistral Large 3 | Qwen2.5 VL 72B Instruct |
|---|---|---|
| OSWorld | — | 8.8 |
Category Winners
general
17.1
multimodal
Q
Qwen2.5 VL 72B Instruct25.0
agent
Q
Qwen2.5 VL 72B Instruct3.8