Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | GLM-4.6 | o4-mini |
|---|---|---|
| GPQA Diamond | 81 | 81.4 |
| Humanity's Last Exam | 17.2 | 14.7 |
math
| Benchmark | GLM-4.6 | o4-mini |
|---|---|---|
| AIME 2025 | 93.9 | 92.7 |
coding
| Benchmark | GLM-4.6 | o4-mini |
|---|---|---|
| SWE-bench Verified | 68 | 68.1 |
| TerminalBench | 40.5 | — |
agent
| Benchmark | GLM-4.6 | o4-mini |
|---|---|---|
| BrowseComp | 45.1 | 51.5 |
| TAU-Bench Retail | — | 71.8 |
multimodal
| Benchmark | GLM-4.6 | o4-mini |
|---|---|---|
| MMMU | — | 81.6 |
| CharXiv Reasoning | — | 72 |
Category Winners
coding
Z
GLM-4.663.6
math
Z
GLM-4.676.1
reasoning
Z
GLM-4.647.8
multimodal
62.8
agent
43.9