Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | GPT-4.1 mini | Qwen2.5 VL 72B Instruct |
|---|---|---|
| GPQA Diamond | 65 | — |
| Humanity's Last Exam | 3.7 | — |
math
| Benchmark | GPT-4.1 mini | Qwen2.5 VL 72B Instruct |
|---|---|---|
| AIME 2025 | 40.2 | — |
coding
| Benchmark | GPT-4.1 mini | Qwen2.5 VL 72B Instruct |
|---|---|---|
| SWE-bench Verified | 23.6 | — |
multimodal
| Benchmark | GPT-4.1 mini | Qwen2.5 VL 72B Instruct |
|---|---|---|
| MMMU | 72.7 | 70.2 |
| CharXiv Reasoning | 56.8 | — |
| MMMU-Pro | — | 51.1 |
| ScreenSpot Pro | — | 43.6 |
general
| Benchmark | GPT-4.1 mini | Qwen2.5 VL 72B Instruct |
|---|---|---|
| MMMLU | 78.5 | — |
agent
| Benchmark | GPT-4.1 mini | Qwen2.5 VL 72B Instruct |
|---|---|---|
| TAU-Bench Retail | 55.8 | — |
| OSWorld | — | 8.8 |
Category Winners
coding
0.6
math
3.2
reasoning
26.6
general
31.1
multimodal
39.3
agent
18.1