Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Gemini 3.1 Flash-Lite | Qwen 2.5 72B |
|---|---|---|
| GPQA Diamond | 86.9 | 49 |
| Humanity's Last Exam | 16 | — |
| BigBench-Hard | — | 83.5 |
| ARC-Challenge | — | 95.3 |
multimodal
| Benchmark | Gemini 3.1 Flash-Lite | Qwen 2.5 72B |
|---|---|---|
| MMMU-Pro | 76.8 | — |
| CharXiv Reasoning | 73.2 | — |
general
safety
| Benchmark | Gemini 3.1 Flash-Lite | Qwen 2.5 72B |
|---|---|---|
| TruthfulQA | — | 62.5 |
language
| Benchmark | Gemini 3.1 Flash-Lite | Qwen 2.5 72B |
|---|---|---|
| HellaSwag | — | 94 |
| WinoGrande | — | 84.8 |
coding
| Benchmark | Gemini 3.1 Flash-Lite | Qwen 2.5 72B |
|---|---|---|
| HumanEval | — | 86.6 |
Category Winners
coding
59.0
math
78.5
reasoning
50.6
general
65.0
language
65.7
multimodal
58.3
safety
41.8