Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
coding
| Benchmark | Claude Opus 4 | Gemini 3.1 Pro |
|---|---|---|
| SWE-bench Verified | 55.8 | 80.6 |
| HumanEval | 95.1 | — |
| SWE-Bench Pro | — | 54.2 |
| SciCode | — | 59 |
general
| Benchmark | Claude Opus 4 | Gemini 3.1 Pro |
|---|---|---|
| IFEval | 91.5 | — |
| MMLU-Pro | 83.5 | — |
| Chatbot Arena Elo | 1365 | — |
| MMMLU | — | 92.6 |
| MRCR v2 |
reasoning
| Benchmark | Claude Opus 4 | Gemini 3.1 Pro |
|---|---|---|
| GPQA Diamond | 74.8 | 94.3 |
| BigBench-Hard | 93.2 | — |
| Humanity's Last Exam | — | 51.4 |
| ARC-AGI v2 | — | 77.1 |
agent
| Benchmark | Claude Opus 4 | Gemini 3.1 Pro |
|---|---|---|
| BrowseComp | — | 85.9 |
| MCP Atlas | — | 69.2 |
| APEX Agents | — | 33.5 |
Category Winners
coding
68.1
math
75.5
reasoning
89.3
general
96.0
multimodal
81.8
agent
68.9