Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
safety
| Benchmark | GPT-4o mini | Qwen3 235B A22B |
|---|---|---|
| TruthfulQA | 65.8 | — |
reasoning
| Benchmark | GPT-4o mini | Qwen3 235B A22B |
|---|---|---|
| BigBench-Hard | 78.5 | — |
| ARC-Challenge | 93.1 | — |
| GPQA Diamond | 40.2 | 47.5 |
coding
| Benchmark | GPT-4o mini | Qwen3 235B A22B |
|---|---|---|
| HumanEval | 87.2 | — |
general
| Benchmark | GPT-4o mini | Qwen3 235B A22B |
|---|---|---|
| MMLU-Pro | 63.2 | — |
| Chatbot Arena Elo | 1219 | — |
| IFEval | 80.1 | — |
| MMMLU | — | 86.7 |
language
| Benchmark | GPT-4o mini | Qwen3 235B A22B |
|---|---|---|
| HellaSwag | 91.2 | — |
| WinoGrande | 82.3 | — |
Category Winners
coding
61.6
math
Q
Qwen3 235B A22B54.5
reasoning
19.2
general
Q
Qwen3 235B A22B63.7
language
35.0
safety
62.9