Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Muse Spark | Nova 2 Pro |
|---|---|---|
| GPQA Diamond | 89.5 | 81.4 |
| Humanity's Last Exam | 58.4 | — |
| ARC-AGI v2 | 42.5 | — |
coding
| Benchmark | Muse Spark | Nova 2 Pro |
|---|---|---|
| SWE-bench Verified | 77.4 | 70 |
| SWE-Bench Pro | 52.4 | — |
| TerminalBench | — | 41.3 |
multimodal
| Benchmark | Muse Spark | Nova 2 Pro |
|---|---|---|
| MMMU-Pro | 80.4 | 63.5 |
| CharXiv Reasoning | 86.4 | — |
| ScreenSpot Pro | 84.1 | — |
math
| Benchmark | Muse Spark | Nova 2 Pro |
|---|---|---|
| AIME 2025 | — | 92.3 |
Category Winners
coding
Nova 2 Pro67.3
math
Nova 2 Pro73.5
reasoning
Nova 2 Pro76.8
multimodal
M
Muse Spark85.1