Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Gemini 2.5 Pro | MiniMax M2.1 |
|---|---|---|
| GPQA Diamond | 83 | 81 |
| Humanity's Last Exam | 17.8 | 22 |
| ARC-AGI v2 | 4.9 | — |
math
| Benchmark | Gemini 2.5 Pro | MiniMax M2.1 |
|---|---|---|
| AIME 2025 | 83 | 81 |
coding
| Benchmark | Gemini 2.5 Pro | MiniMax M2.1 |
|---|---|---|
| SWE-bench Verified | 63.2 | 67 |
| TerminalBench | — | 47.9 |
| SciCode | — | 39 |
multimodal
| Benchmark | Gemini 2.5 Pro | MiniMax M2.1 |
|---|---|---|
| MMMU | 79.6 | — |
general
| Benchmark | Gemini 2.5 Pro | MiniMax M2.1 |
|---|---|---|
| SimpleQA | 50.8 | — |
agent
| Benchmark | Gemini 2.5 Pro | MiniMax M2.1 |
|---|---|---|
| BrowseComp | — | 62 |
| Toolathlon | — | 43.5 |
Category Winners
coding
55.4
math
57.4
reasoning
54.0
general
68.7
multimodal
80.3
agent
38.2