Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
coding
| Benchmark | Claude Fable 5 | Claude Opus 5 |
|---|---|---|
| SWE-bench Verified | 95 | — |
| SWE-Bench Pro | 80 | — |
reasoning
| Benchmark | Claude Fable 5 | Claude Opus 5 |
|---|---|---|
| Humanity's Last Exam | 64.5 | 64.7 |
agent
| Benchmark | Claude Fable 5 | Claude Opus 5 |
|---|---|---|
| BrowseComp | — | 90.8 |
Category Winners
coding
97.5
reasoning
96.5
agent
89.8