Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Claude 3.5 Sonnet | GPT-4.1 |
|---|---|---|
| GPQA Diamond | 59.4 | 66.3 |
| Humanity's Last Exam | — | 5.4 |
math
| Benchmark | Claude 3.5 Sonnet | GPT-4.1 |
|---|---|---|
| AIME 2025 | — | 46.4 |
coding
| Benchmark | Claude 3.5 Sonnet | GPT-4.1 |
|---|---|---|
| SWE-bench Verified | — | 54.6 |
multimodal
| Benchmark | Claude 3.5 Sonnet | GPT-4.1 |
|---|---|---|
| MMMU | — | 74.8 |
| CharXiv Reasoning | — | 56.7 |
general
| Benchmark | Claude 3.5 Sonnet | GPT-4.1 |
|---|---|---|
| MMMLU | — | 87.3 |
agent
| Benchmark | Claude 3.5 Sonnet | GPT-4.1 |
|---|---|---|
| TAU-Bench Retail | — | 68 |
Category Winners
coding
25.5
math
5.8
reasoning
34.0
general
66.0
multimodal
39.9
agent
50.7