GLM-5.1 FP8: benchmark results
GLM-5.1 evaluated in FP8 quantization. Provider: Zhipu. Released 2026-03-27. Access: Open.
Aggregate Rating 1755 ± 27, rank #109 of 1656 rated models, from 16 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| RewardBench 2 Precise IF | 73.28 | Score (%) | 98 |
| RewardBench 2 Factuality | 84.47 | Accuracy (%) | 97.2 |
| MATH-MC Level 4 | 99.58 | Accuracy (%) | 95.6 |
| RewardBench 2 Math | 89.07 | Score (%) | 93.1 |
| JudgeBench Knowledge | 88.31 | Accuracy (%) | 92.2 |
| MATH-MC Level 5 | 99.69 | Accuracy (%) | 90.4 |
| JudgeBench Coding | 97.62 | Accuracy (%) | 88.2 |
| MATH-MC Level 1 | 98.84 | Accuracy (%) | 85.3 |
| RewardBench 2 Focus | 90.15 | Score (%) | 82.1 |
| GSM-MC | 99.24 | Accuracy (%) | 81.3 |
| MATH-MC Level 3 | 99.1 | Accuracy (%) | 79.4 |
| MATH-MC Level 2 | 98.53 | Accuracy (%) | 71.3 |
Interactive version: theaggregate.ai/model?slug=glm-5-1-fp8 · How It Works · Data refreshed daily, snapshot 2026-10-20.