LongWriter-glm4-9B: benchmark results
zai-org's LongWriter GLM-4 9B, fine-tuned from GLM-4-9B to generate over 10,000 words in a single response. Provider: Zhipu. Released 2024-08-12. Access: Open.
Aggregate Rating 1499 ± 32, rank #778 of 1635 rated models, from 18 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Portuguese LLM - BLUEX | 59.81 | Accuracy (%) | 76.9 |
| Open Portuguese LLM - ENEM | 70.19 | Accuracy (%) | 75.2 |
| Enkrypt AI - Bias Risk | 82.17 | Risk Score | 54.4 |
| Enkrypt AI - Jailbreak Risk | 8.53 | Risk Score | 54.4 |
| Open Portuguese LLM - OAB Exams | 42.41 | Accuracy (%) | 51.6 |
| Enkrypt AI - Insecure Code Risk | 31.11 | Risk Score | 45.1 |
| Enkrypt AI - Risk Score | 33.96 | Risk Score | 41.9 |
| Open Portuguese LLM - ASSIN2 RTE | 84.03 | Macro F1 (%) | 33.8 |
| EduResearchBench - Peer Review | 2.62 | Judge score (1-5; 2 atomic tasks (language quality, logic fl | 28.6 |
| EduResearchBench - Theory and Survey | 1.77 | Judge score (1-5; 5 atomic tasks, Chinese only, mean of Doub | 28.6 |
| Enkrypt AI - Safety Risk | 37.71 | Risk Score | 18 |
| Enkrypt AI - Toxicity Risk | 8.64 | Risk Score | 17.3 |
Interactive version: theaggregate.ai/model?slug=longwriter-glm4-9b · How It Works · Data refreshed daily, snapshot 2026-11-05.