Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
coding
| Benchmark | Claude Opus 4 | Qwen3.8-Flash-Next |
|---|---|---|
| HumanEval | 95.1 | — |
| SWE-bench Verified | 55.8 | — |
| SWE-Bench Pro | — | 62.5 |
reasoning
| Benchmark | Claude Opus 4 | Qwen3.8-Flash-Next |
|---|---|---|
| GPQA Diamond | 74.8 | 91.7 |
| BigBench-Hard | 93.2 | — |
| Humanity's Last Exam | — | 35.9 |
general
| Benchmark | Claude Opus 4 | Qwen3.8-Flash-Next |
|---|---|---|
| Chatbot Arena Elo | 1365 | — |
| MMLU-Pro | 83.5 | — |
| IFEval | 91.5 | — |
multimodal
| Benchmark | Claude Opus 4 | Qwen3.8-Flash-Next |
|---|---|---|
| MMMU | 74.2 | — |
| CharXiv Reasoning | — | 90.6 |
agent
| Benchmark | Claude Opus 4 | Qwen3.8-Flash-Next |
|---|---|---|
| Toolathlon | — | 73.5 |
Category Winners
coding
Q
Qwen3.8-Flash-Next67.2
math
75.5
reasoning
79.3
general
96.0
multimodal
Q
Qwen3.8-Flash-Next86.6
agent
Q
Qwen3.8-Flash-Next92.7