GLM-4.6V (Non-reasoning): benchmark results
GLM-4.6V evaluated with reasoning disabled. Provider: Zhipu. Released 2025-12-08. Access: Open.
Aggregate Rating 1545 ± 14, rank #903 of 2168 rated models, from 101 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| DocFormBench (Docx-Skill) | 25.57 | Format accuracy (%) on DocFormBench: share of required forma | 100 |
| MultihopSpatial - 1-Hop Exo-Centric | 80.1 | Multiple-choice accuracy (%) on the 1-hop exo-centric questi | 94.4 |
| MultihopSpatial - Acc@50IoU | 35.2 | Acc@50IoU (%): a prediction counts only when the answer is r | 88.9 |
| UAV-DualCog - Flight Behavior Recognition (Atomic) | 32.5 | Atomic-level behavior accuracy (%; the atomic flight actions | 84.6 |
| MultihopSpatial | 43.2 | Multiple-choice accuracy (%) over all 4,500 MultihopSpatial | 80.6 |
| MultihopSpatial - 2-Hop Exo-Centric | 61.6 | Multiple-choice accuracy (%) on the 2-hop exo-centric questi | 77.8 |
| UGI - Natural Intelligence | 32.31 | NatInt Score | 76.6 |
| MultihopSpatial - 1-Hop Ego-Centric | 32.4 | Multiple-choice accuracy (%) on the 1-hop ego-centric questi | 76.4 |
| MultihopSpatial - 3-Hop Exo-Centric | 46.7 | Multiple-choice accuracy (%) on the 3-hop exo-centric questi | 75 |
| UAV-DualCog - Flight Behavior Recognition (Composite) | 23.2 | Composite-level behavior accuracy (%; 1,024 first-person UAV | 69.2 |
| DocFormBench (MCP-Msoffice) | 16.38 | Format accuracy (%) on DocFormBench: share of required forma | 66.7 |
| AA-Omniscience Hallucination Rate | 66.63 | Hallucination Rate (%) | 64.9 |
Interactive version: theaggregate.ai/model?slug=glm-4-6v-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.