Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | EXAONE 4.5 33B | GPT-5.2 |
|---|---|---|
| GPQA Diamond | 80.5 | 92.4 |
| Humanity's Last Exam | — | 34.5 |
| ARC-AGI v2 | — | 52.9 |
math
| Benchmark | EXAONE 4.5 33B | GPT-5.2 |
|---|---|---|
| AIME 2025 | 92.9 | 100 |
| FrontierMath | — | 40.3 |
multimodal
| Benchmark | EXAONE 4.5 33B | GPT-5.2 |
|---|---|---|
| MMMU | 78.7 | — |
| MMMU-Pro | 68.6 | 79.5 |
| CharXiv Reasoning | — | 82.1 |
| ScreenSpot Pro | — | 86.3 |
coding
| Benchmark | EXAONE 4.5 33B | GPT-5.2 |
|---|---|---|
| SWE-bench Verified | — | 80 |
general
| Benchmark | EXAONE 4.5 33B | GPT-5.2 |
|---|---|---|
| MMMLU | — | 89.6 |
agent
| Benchmark | EXAONE 4.5 33B | GPT-5.2 |
|---|---|---|
| BrowseComp | — | 65.8 |
| Toolathlon | — | 46.3 |
| MCP Atlas | — | 60.6 |
Category Winners
coding
80.8
math
L
EXAONE 4.5 33B74.6
reasoning
L
EXAONE 4.5 33B73.4
general
74.6
multimodal
80.0
agent
34.3