Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude Opus 4.5 | Step-3.5-Flash |
|---|---|---|
| GPQA Diamond | 87 | — |
| ARC-AGI v2 | 37.6 | — |
coding
| Benchmark | Claude Opus 4.5 | Step-3.5-Flash |
|---|---|---|
| SWE-bench Verified | 80.9 | 74.4 |
general
| Benchmark | Claude Opus 4.5 | Step-3.5-Flash |
|---|---|---|
| MMMLU | 90.8 | — |
agent
| Benchmark | Claude Opus 4.5 | Step-3.5-Flash |
|---|---|---|
| OSWorld | 66.3 | — |
| MCP Atlas | 62.3 | — |
| BrowseComp | — | 69 |
math
| Benchmark | Claude Opus 4.5 | Step-3.5-Flash |
|---|---|---|
| AIME 2025 | — | 97.3 |
Category Winners
coding
82.5
math
S
Step-3.5-Flash80.7
reasoning
61.5
general
78.0
agent
S
Step-3.5-Flash58.6