GLM-5.3: benchmark results
Provider: Zhipu. Released 2026-08-14. Access: API.
Unified ELO 1715 ± 1, rank #36 of 1392 rated models, from 134 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Featherbench | 100 | Pass Rate (%) | 100 |
| Humanity's Last Exam (Self-Reported, With Tools) | 62.5 | Accuracy (%) | 100 |
| LLM Stats (PostTrainBench) | 39.8 | Score (%) | 100 |
| LLM Stats (SWE-Marathon) | 42.5 | Score (%) | 100 |
| SWE-Milestone | 58.75 | Milestone Score (%) | 100 |
| SecIT Bench (Pydantic AI) | 81.44 | Accuracy (%) | 100 |
| Z.ai GLM-5.3 Launch - GDPval-AA v2 | 1769 | ELO | 100 |
| LiveBench Theory of Mind | 88.46 | Score | 99.1 |
| EQ-Bench Longform Writing | 81.8 | Writing Score (0-100) | 98.1 |
| LLM Stats Score | 53.6 | LLM Stats Score (conservative rating) | 97.7 |
| OpenRouter Tau2-Bench Airline | 80 | Accuracy (%) | 97.5 |
| BenchmarkList ECI | 148.78 | Capability Index (ECI) | 95.1 |
Interactive version: theaggregate.ai/model?slug=glm-5-3 · How It Works · Data refreshed daily, snapshot 2026-09-05.