GLM-4.6 (Non-reasoning): benchmark results
GLM-4.6 evaluated with reasoning disabled. Provider: Zhipu. Released 2025-09-30. Access: Open.
Aggregate Rating 1632 ± 14, rank #594 of 2168 rated models, from 82 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AgentNoiseBench (Search, Tool Noise) | 39 | Accuracy on 500 sampled 2WikiMultiHopQA and HotpotQA questio | 100 |
| DocFormBench (DocFormFlow) - Hallucinated Formatting | 1.24 | Hallucinated formatting rate (%) on DocFormBench: share of a | 100 |
| WuYuEval - Foundation Module - Accuracy | 92.31 | Accuracy (%) | 93.8 |
| WuYuEval - Foundation Module - Macro-F1 | 93.54 | Macro-F1 (%) | 93.8 |
| DocFormBench (DocFormFlow) | 75.72 | Format accuracy (%) on DocFormBench: share of required forma | 92.9 |
| BFI-Adapt | 32.2 | BFI-Adapt composite (%; mean over the 27 event-trait pairs w | 92.3 |
| AA Omniscience - Software Engineering (SWE) - Julia | 32 | Accuracy (%) | 90.1 |
| UGI Leaderboard | 49.83 | UGI Score | 88.6 |
| UGI - Writing | 50.01 | Writing Score | 86.6 |
| AA Omniscience - Software Engineering (SWE) - Java | 26 | Accuracy (%) | 86 |
| UGI - Natural Intelligence | 41.7 | NatInt Score | 85.7 |
| DocFormBench (MCP-Msoffice) | 18.03 | Format accuracy (%) on DocFormBench: share of required forma | 83.3 |
Interactive version: theaggregate.ai/model?slug=glm-4-6-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-10-20.