Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | GPT-5 Medium | Qwen3.8 Flash |
|---|---|---|
| GPQA Diamond | 88.1 | 91.7 |
| Humanity's Last Exam | — | 35.9 |
math
| Benchmark | GPT-5 Medium | Qwen3.8 Flash |
|---|---|---|
| AIME 2025 | 88.9 | — |
coding
| Benchmark | GPT-5 Medium | Qwen3.8 Flash |
|---|---|---|
| SWE-Bench Pro | — | 62.5 |
agent
| Benchmark | GPT-5 Medium | Qwen3.8 Flash |
|---|---|---|
| Toolathlon | — | 73.5 |
multimodal
| Benchmark | GPT-5 Medium | Qwen3.8 Flash |
|---|---|---|
| CharXiv Reasoning | — | 90.6 |
Category Winners
coding
Q
Qwen3.8 Flash72.9
math
68.1
reasoning
84.8
multimodal
Q
Qwen3.8 Flash85.9
agent
Q
Qwen3.8 Flash93.1