GLM-4.1V-9B-Base: benchmark results
Provider: Zhipu. Released 2025-07-01. Access: Open.
Aggregate Rating 1577 ± 27, rank #1302 of 10916 rated models, from 52 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| DatBench - Diagrams & Tables | 66.4 | Accuracy (%) | 100 |
| DatBench - Spatial Reasoning | 36.8 | Accuracy (%) | 100 |
| NutriBench-Kitchen - Ingredient Entry (Easy) | 38 | Score (%; the ingredient entry task family, easy regime: sin | 90.9 |
| AeroGround - Relative Orientation Determination | 34.4 | Accuracy (%; Spatial Understanding: select the egocentric qu | 90 |
| AeroGround - Trajectory Reconstruction Selection | 33 | Accuracy (%; Spatial Reasoning: select the aerial trajectory | 86.7 |
| DatBench - Counting | 92.4 | Accuracy (%) | 86.5 |
| DatBench - Chart Understanding | 66.3 | Accuracy (%) | 84.6 |
| DatBench - Scene OCR | 73.6 | Accuracy (%) | 84.6 |
| DatBench - General VQA | 54.1 | Accuracy (%) | 78.8 |
| AeroGround - Region Counting | 37.5 | Accuracy (%; Spatial Understanding: count category instances | 76.7 |
| AeroGround - Occluded Target Judgment | 40.3 | Accuracy (%; Spatial Reasoning: decide whether a target is v | 73.3 |
| NutriBench-Kitchen - Long-Term Planning (Easy) | 90.7 | Score (%; the long-term planning task family, easy regime: s | 72.7 |
Interactive version: theaggregate.ai/model?slug=glm-4-1v-9b-base · How It Works · Data refreshed daily, snapshot 2026-11-08.