Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude 3 Opus | Gemini 2.5 Flash-Lite |
|---|---|---|
| GPQA Diamond | 50.4 | 64.6 |
| Humanity's Last Exam | — | 5.1 |
math
| Benchmark | Claude 3 Opus | Gemini 2.5 Flash-Lite |
|---|---|---|
| AIME 2025 | — | 49.8 |
coding
| Benchmark | Claude 3 Opus | Gemini 2.5 Flash-Lite |
|---|---|---|
| SWE-bench Verified | — | 31.6 |
multimodal
| Benchmark | Claude 3 Opus | Gemini 2.5 Flash-Lite |
|---|---|---|
| MMMU | — | 72.9 |
general
| Benchmark | Claude 3 Opus | Gemini 2.5 Flash-Lite |
|---|---|---|
| SimpleQA | — | 10.7 |
Category Winners
coding
2.2
math
8.1
reasoning
26.8
general
14.4
multimodal
64.8