Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | GPT-4.1 | Grok-2 |
|---|---|---|
| GPQA Diamond | 66.3 | 56 |
| Humanity's Last Exam | 5.4 | — |
math
| Benchmark | GPT-4.1 | Grok-2 |
|---|---|---|
| AIME 2025 | 46.4 | — |
coding
| Benchmark | GPT-4.1 | Grok-2 |
|---|---|---|
| SWE-bench Verified | 54.6 | — |
multimodal
| Benchmark | GPT-4.1 | Grok-2 |
|---|---|---|
| MMMU | 74.8 | 66.1 |
| CharXiv Reasoning | 56.7 | — |
general
| Benchmark | GPT-4.1 | Grok-2 |
|---|---|---|
| MMMLU | 87.3 | — |
agent
| Benchmark | GPT-4.1 | Grok-2 |
|---|---|---|
| TAU-Bench Retail | 68 | — |
Category Winners
coding
26.2
math
6.0
reasoning
X
Grok-229.6
general
66.0
multimodal
X
Grok-246.0
agent
50.7