Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
coding
| Benchmark | Claude 3.5 Sonnet | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| SWE-bench Verified | 49 | — |
| HumanEval | 92 | — |
language
| Benchmark | Claude 3.5 Sonnet | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| HellaSwag | 89 | — |
| WinoGrande | 83.2 | — |
reasoning
| Benchmark | Claude 3.5 Sonnet | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| ARC-Challenge | 96.7 | — |
| GPQA Diamond | 65 | 77.5 |
| BigBench-Hard | 89.4 | — |
multimodal
| Benchmark | Claude 3.5 Sonnet | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| MMMU | 68.3 | — |
safety
| Benchmark | Claude 3.5 Sonnet | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| TruthfulQA | 74.1 | — |
general
| Benchmark | Claude 3.5 Sonnet | Qwen3-235B-A22B-Instruct-2507 |
|---|---|---|
| Chatbot Arena Elo | 1271 | — |
| MMLU-Pro | 78 | — |
| IFEval | 88 | — |
| SimpleQA | — | 54.3 |
Category Winners
coding
47.8
math
58.6
reasoning
Q
Qwen3-235B-A22B-Instruct-250768.6
general
80.5
language
31.6
multimodal
50.6
safety
95.3