Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | DiffusionGemma 26B-A4B | GLM-4.5-Air |
|---|---|---|
| GPQA Diamond | 73.2 | 75 |
| Humanity's Last Exam | 11.9 | 10.6 |
multimodal
| Benchmark | DiffusionGemma 26B-A4B | GLM-4.5-Air |
|---|---|---|
| MMMU-Pro | 54.3 | — |
general
| Benchmark | DiffusionGemma 26B-A4B | GLM-4.5-Air |
|---|---|---|
| MMMLU | 81.5 | — |
coding
| Benchmark | DiffusionGemma 26B-A4B | GLM-4.5-Air |
|---|---|---|
| SWE-bench Verified | — | 57.6 |
| TerminalBench | — | 30 |
| SciCode | — | 37.3 |
agent
| Benchmark | DiffusionGemma 26B-A4B | GLM-4.5-Air |
|---|---|---|
| BrowseComp | — | 21.3 |
| TAU-Bench Retail | — | 77.9 |
Category Winners
coding
Z
GLM-4.5-Air29.5
reasoning
Z
GLM-4.5-Air37.6
general
42.6
multimodal
15.4
agent
Z
GLM-4.5-Air39.6