Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude Opus 4.7 | Hy4 preview |
|---|---|---|
| GPQA Diamond | 94.2 | 92.3 |
| Humanity's Last Exam | 54.7 | — |
coding
| Benchmark | Claude Opus 4.7 | Hy4 preview |
|---|---|---|
| SWE-bench Verified | 87.6 | — |
| SWE-Bench Pro | 64.3 | 65.7 |
general
| Benchmark | Claude Opus 4.7 | Hy4 preview |
|---|---|---|
| MMMLU | 91.5 | — |
agent
| Benchmark | Claude Opus 4.7 | Hy4 preview |
|---|---|---|
| BrowseComp | 79.3 | — |
| MCP Atlas | 77.3 | 83.7 |
| Toolathlon | — | 74.1 |
| APEX Agents | — | 37.1 |
multimodal
| Benchmark | Claude Opus 4.7 | Hy4 preview |
|---|---|---|
| CharXiv Reasoning | 91 | — |
Category Winners
coding
82.3
reasoning
90.2
general
80.6
multimodal
87.2
agent
T
Hy4 preview77.3