Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude Mythos Preview | Kimi K3 |
|---|---|---|
| GPQA Diamond | 94.6 | 93.5 |
| Humanity's Last Exam | 64.7 | 56 |
coding
| Benchmark | Claude Mythos Preview | Kimi K3 |
|---|---|---|
| SWE-bench Verified | 93.9 | — |
| SWE-Bench Pro | 77.8 | — |
general
| Benchmark | Claude Mythos Preview | Kimi K3 |
|---|---|---|
| MMMLU | 92.7 | — |
agent
| Benchmark | Claude Mythos Preview | Kimi K3 |
|---|---|---|
| BrowseComp | 86.9 | 91.2 |
| Toolathlon | — | 73.2 |
| MCP Atlas | — | 84.2 |
| APEX Agents | — | 37.6 |
multimodal
| Benchmark | Claude Mythos Preview | Kimi K3 |
|---|---|---|
| CharXiv Reasoning | 93.2 | 91.3 |
| MMMU-Pro | — | 81.6 |
Category Winners
coding
97.7
reasoning
94.6
general
83.3
multimodal
91.8
agent
M
Kimi K393.2