Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
language
| Benchmark | Qwen 2.5 72B | Qwen3-235B-A22B-Thinking-2507 |
|---|---|---|
| WinoGrande | 84.8 | — |
| HellaSwag | 94 | — |
general
| Benchmark | Qwen 2.5 72B | Qwen3-235B-A22B-Thinking-2507 |
|---|---|---|
| IFEval | 84.2 | — |
| Chatbot Arena Elo | 1261 | — |
| MMLU-Pro | 71.1 | — |
coding
| Benchmark | Qwen 2.5 72B | Qwen3-235B-A22B-Thinking-2507 |
|---|---|---|
| HumanEval | 86.6 | — |
reasoning
| Benchmark | Qwen 2.5 72B | Qwen3-235B-A22B-Thinking-2507 |
|---|---|---|
| ARC-Challenge | 95.3 | — |
| BigBench-Hard | 83.5 | — |
| GPQA Diamond | 49 | 81.1 |
| Humanity's Last Exam | — | 18.2 |
safety
| Benchmark | Qwen 2.5 72B | Qwen3-235B-A22B-Thinking-2507 |
|---|---|---|
| TruthfulQA | 62.5 | — |
agent
| Benchmark | Qwen 2.5 72B | Qwen3-235B-A22B-Thinking-2507 |
|---|---|---|
| TAU-Bench Retail | — | 67.8 |
Category Winners
coding
59.0
math
78.5
reasoning
Q
Qwen3-235B-A22B-Thinking-250748.6
general
63.7
language
65.7
safety
41.8
agent
Q
Qwen3-235B-A22B-Thinking-250750.1