Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| Humanity's Last Exam | 64.5 | — |
coding
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| SWE-bench Verified | 95 | — |
| SWE-Bench Pro | 80 | — |
agent
| Benchmark | Claude Fable 5 | Grok 4.6 |
|---|---|---|
| APEX Agents | — | 57.5 |
Category Winners
coding
98.1
reasoning
96.6
agent
X
Grok 4.699.3