Neura Intelligence Index
| Benchmark | Kimi K3 | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 93.5 | 92.6 |
| Humanity's Last Exam | 56 | 43.6 |
| Benchmark | Kimi K3 | Qwen3.8 Max |
|---|---|---|
| MMMU-Pro | 81.6 | 82.3 |
| CharXiv Reasoning | 91.3 | — |
| ScreenSpot Pro | — | 84.5 |
| Benchmark | Kimi K3 | Qwen3.8 Max |
|---|---|---|
| BrowseComp | 91.2 | — |
| Toolathlon | 73.2 | 72.5 |
| MCP Atlas | 84.2 | — |
| APEX Agents | 37.6 | — |
| Benchmark | Kimi K3 | Qwen3.8 Max |
|---|---|---|
| SWE-Bench Pro | — | 67.7 |
| Benchmark | Kimi K3 | Qwen3.8 Max |
|---|---|---|
| MRCR v2 | — | 92.9 |