GLM-4V: benchmark results
Provider: Zhipu. Released 2024-06-05.
Aggregate Rating 1448 ± 37, rank #3864 of 10916 rated models, from 207 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM MMBench V1.1 CN - Image Emotion | 87.8 | Accuracy (%) | 98.6 |
| OpenVLM ScienceQA Val | 98.6 | Accuracy (%) | 98.2 |
| OpenVLM MMBench V1.1 CN - Identity Reasoning | 100 | Accuracy (%) | 97.9 |
| OpenVLM POPE | 89.9 | Overall (%) | 97.2 |
| OpenVLM MMBench V1.1 CN - Celebrity Recognition | 93.9 | Accuracy (%) | 96.3 |
| OpenVLM LLaVA-Bench - Detail | 91.4 | Score | 94 |
| OpenVLM MMBench V1.1 CN - Action Recognition | 92.3 | Accuracy (%) | 92.6 |
| OpenVLM MMBench V1.1 EN - Celebrity Recognition | 91.8 | Accuracy (%) | 92 |
| OpenVLM MMBench V1.1 EN - Identity Reasoning | 98.7 | Accuracy (%) | 91.5 |
| OpenVLM ScienceQA Test | 97.4 | Accuracy (%) | 91.5 |
| OpenVLM MMBench V1.1 EN - Action Recognition | 92.3 | Accuracy (%) | 90.5 |
| OpenVLM MME - Existence | 200 | Score | 89.7 |
Interactive version: theaggregate.ai/model?slug=glm-4v · How It Works · Data refreshed daily, snapshot 2026-11-08.