Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Gemma 4 31B | Qwen3-Next-80B-A3B-Thinking |
|---|---|---|
| GPQA Diamond | 84.3 | 77.2 |
| Humanity's Last Exam | 26.5 | — |
multimodal
| Benchmark | Gemma 4 31B | Qwen3-Next-80B-A3B-Thinking |
|---|---|---|
| MMMU-Pro | 76.9 | — |
math
| Benchmark | Gemma 4 31B | Qwen3-Next-80B-A3B-Thinking |
|---|---|---|
| AIME 2025 | — | 87.8 |
agent
| Benchmark | Gemma 4 31B | Qwen3-Next-80B-A3B-Thinking |
|---|---|---|
| TAU-Bench Retail | — | 69.6 |
Category Winners
math
Q
Qwen3-Next-80B-A3B-Thinking66.1
reasoning
Q
Qwen3-Next-80B-A3B-Thinking68.1
general
71.5
multimodal
74.1
agent
Q
Qwen3-Next-80B-A3B-Thinking55.5