GLM-4.6 FP8: benchmark results
GLM-4.6 evaluated in FP8 quantization. Provider: Zhipu. Released 2025-09-30. Access: Open.
Aggregate Rating 1667 ± 27, rank #278 of 1656 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| RewardBench 2 Math | 90.44 | Score (%) | 97.8 |
| RewardBench 2 Precise IF | 64.06 | Score (%) | 91.9 |
| MATH-MC Level 5 | 99.69 | Accuracy (%) | 90.4 |
| RewardBench 2 Factuality | 77.95 | Accuracy (%) | 88.3 |
| RewardBench 2 Focus | 90.91 | Score (%) | 86.5 |
| MolecularIQ - Generation | 22 | Mean Accuracy (%) | 78.4 |
| MolecularIQ - Indexing | 11.3 | Mean Accuracy (%) | 75.7 |
| MolecularIQ - Overall | 16.2 | Mean Accuracy (%) | 75.7 |
| MATH-MC Level 1 | 98.6 | Accuracy (%) | 75 |
| MATH-MC Level 4 | 99.33 | Accuracy (%) | 75 |
| RewardBench 2 Safety | 87.11 | Score (%) | 71.4 |
| MATH-MC Level 2 | 98.53 | Accuracy (%) | 71.3 |
Interactive version: theaggregate.ai/model?slug=glm-4-6-fp8 · How It Works · Data refreshed daily, snapshot 2026-10-20.