Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude 3.5 Sonnet | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 67.2 | — |
coding
| Benchmark | Claude 3.5 Sonnet | Qwen3 32B |
|---|---|---|
| SWE-bench Verified | 49 | — |
multimodal
| Benchmark | Claude 3.5 Sonnet | Qwen3 32B |
|---|---|---|
| MMMU | 68.3 | — |
agent
| Benchmark | Claude 3.5 Sonnet | Qwen3 32B |
|---|---|---|
| TAU-Bench Retail | 69.2 | — |
math
| Benchmark | Claude 3.5 Sonnet | Qwen3 32B |
|---|---|---|
| AIME 2025 | — | 72.9 |
Category Winners
coding
16.5
math
Q
Qwen3 32B38.4
reasoning
51.0
multimodal
52.2
agent
54.3