Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Gemini 2.0 Flash | Qwen3 VL 30B A3B Thinking |
|---|---|---|
| GPQA Diamond | 62.1 | 74.4 |
multimodal
| Benchmark | Gemini 2.0 Flash | Qwen3 VL 30B A3B Thinking |
|---|---|---|
| MMMU | 70.7 | — |
| MMMU-Pro | — | 63 |
| CharXiv Reasoning | — | 56.6 |
| ScreenSpot Pro | — | 57.3 |
math
| Benchmark | Gemini 2.0 Flash | Qwen3 VL 30B A3B Thinking |
|---|---|---|
| AIME 2025 | — | 83.1 |
general
| Benchmark | Gemini 2.0 Flash | Qwen3 VL 30B A3B Thinking |
|---|---|---|
| SimpleQA | — | 23.9 |
agent
| Benchmark | Gemini 2.0 Flash | Qwen3 VL 30B A3B Thinking |
|---|---|---|
| OSWorld | — | 30.6 |
Category Winners
math
Q
Qwen3 VL 30B A3B Thinking57.5
reasoning
Q
Qwen3 VL 30B A3B Thinking65.0
general
Q
Qwen3 VL 30B A3B Thinking29.1
multimodal
58.9
agent
Q
Qwen3 VL 30B A3B Thinking23.9