Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
coding
| Benchmark | GPT-5.3 Codex | Qwen3.6-35B-A3B |
|---|---|---|
| SWE-Bench Pro | 56.8 | 49.5 |
| SWE-bench Verified | — | 73.4 |
reasoning
| Benchmark | GPT-5.3 Codex | Qwen3.6-35B-A3B |
|---|---|---|
| GPQA Diamond | — | 86 |
| Humanity's Last Exam | — | 21.4 |
multimodal
| Benchmark | GPT-5.3 Codex | Qwen3.6-35B-A3B |
|---|---|---|
| MMMU | — | 81.7 |
| MMMU-Pro | — | 75.3 |
| CharXiv Reasoning | — | 78 |
agent
| Benchmark | GPT-5.3 Codex | Qwen3.6-35B-A3B |
|---|---|---|
| Toolathlon | — | 26.9 |
| MCP Atlas | — | 62.8 |
Category Winners
coding
48.9
reasoning
Q
Qwen3.6-35B-A3B55.0
multimodal
Q
Qwen3.6-35B-A3B70.8
agent
Q
Qwen3.6-35B-A3B13.3