Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude Opus 4 | MiMo-V2-Flash |
|---|---|---|
| GPQA Diamond | 79.6 | 83.7 |
| ARC-AGI v2 | 8.6 | — |
| Humanity's Last Exam | — | 22.1 |
math
| Benchmark | Claude Opus 4 | MiMo-V2-Flash |
|---|---|---|
| AIME 2025 | 75.5 | 94.1 |
coding
| Benchmark | Claude Opus 4 | MiMo-V2-Flash |
|---|---|---|
| SWE-bench Verified | 72.5 | 73.4 |
| TerminalBench | 39.2 | 30.5 |
general
| Benchmark | Claude Opus 4 | MiMo-V2-Flash |
|---|---|---|
| MMMLU | 88.8 | — |
agent
| Benchmark | Claude Opus 4 | MiMo-V2-Flash |
|---|---|---|
| TAU-Bench Retail | 81.4 | — |
| BrowseComp | — | 58.3 |
Category Winners
coding
66.6
math
X
MiMo-V2-Flash76.4
reasoning
X
MiMo-V2-Flash54.1
general
71.5
agent
85.0