GLM-5.3 (Low): benchmark results
Provider: Zhipu. Released 2026-08-14. Access: Open.
Aggregate Rating 1740 ± 8, rank #274 of 2201 rated models, from 191 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| DecisionPivot - Core Family FM | 96 | Pair Accuracy (%) | 100 |
| DecisionPivot - Core Family TV | 97 | Pair Accuracy (%) | 100 |
| NTBench - NQ (RAG) Fixed Target ASR@1 | 0 | Attack success rate at 1 (%) | 100 |
| NTBench - NQ (RAG) Fixed Target ASR@3 | 0 | Attack success rate at 3 (%) | 100 |
| DecisionPivot - Core Accuracy | 91.7 | Score (%) | 98.1 |
| DecisionPivot - Core Level 3 | 75.6 | Score (%) | 98.1 |
| DecisionPivot - Core Pair Accuracy | 82.2 | Score (%) | 98.1 |
| LegalScope - Case Automatic | 83.1 | Score (%) | 92.6 |
| LegalScope - Argument Validity | 83.2 | Score (%) | 88.9 |
| LegalScope - Citation Relevance | 79.1 | Score (%) | 88.9 |
| DecisionPivot - Core Family LR | 69.7 | Pair Accuracy (%) | 87.5 |
| DecisionPivot - Core Family RR | 78.7 | Pair Accuracy (%) | 87.5 |
Interactive version: theaggregate.ai/model?slug=glm-5-3-low · How It Works · Data refreshed daily, snapshot 2026-11-12.