Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
language
| Benchmark | Mistral Large 2 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| HellaSwag | 93.6 | — |
general
| Benchmark | Mistral Large 2 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| MMLU-Pro | 69.4 | — |
| IFEval | 82.8 | — |
| Chatbot Arena Elo | 1250 | — |
| SimpleQA | — | 51.9 |
reasoning
| Benchmark | Mistral Large 2 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| ARC-Challenge | 94 | — |
| GPQA Diamond | 48.2 | — |
safety
| Benchmark | Mistral Large 2 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| TruthfulQA | 60.5 | — |
coding
| Benchmark | Mistral Large 2 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| HumanEval | 92 | — |
multimodal
| Benchmark | Mistral Large 2 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| MMMU-Pro | — | 68.1 |
| CharXiv Reasoning | — | 62.1 |
| ScreenSpot Pro | — | 62 |
agent
| Benchmark | Mistral Large 2 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| OSWorld | — | 66.7 |
Category Winners
coding
80.3
math
46.2
reasoning
35.6
general
Q
Qwen3 VL 235B A22B Instruct70.6
language
66.8
multimodal
Q
Qwen3 VL 235B A22B Instruct36.5
safety
29.8
agent
Q
Qwen3 VL 235B A22B Instruct