Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Gemini 3 Pro | Qwen3.7-Plus |
|---|---|---|
| GPQA Diamond | 91.9 | 90.3 |
| Humanity's Last Exam | 45.8 | 34.7 |
| ARC-AGI v2 | 31.1 | — |
math
| Benchmark | Gemini 3 Pro | Qwen3.7-Plus |
|---|---|---|
| AIME 2025 | 100 | — |
coding
| Benchmark | Gemini 3 Pro | Qwen3.7-Plus |
|---|---|---|
| SWE-bench Verified | 76.2 | 77.7 |
| SWE-Bench Pro | — | 57.6 |
| SciCode | — | 51.3 |
multimodal
| Benchmark | Gemini 3 Pro | Qwen3.7-Plus |
|---|---|---|
| MMMU-Pro | 81 | 79 |
| CharXiv Reasoning | 81.4 | 85.9 |
| ScreenSpot Pro | 72.7 | 79 |
agent
| Benchmark | Gemini 3 Pro | Qwen3.7-Plus |
|---|---|---|
| MCP Atlas | — | 73.2 |
Category Winners
coding
74.3
math
84.0
reasoning
Q
Qwen3.7-Plus73.8
general
Q
Qwen3.7-Plus72.2
multimodal
Q
Qwen3.7-Plus82.7
agent
Q
Qwen3.7-Plus59.5