Zhipu AI (Z.ai)
GLM-4.5 benchmark results
5 results on 4 benchmarks from 4 sources.
Open weights Released: Aug 3, 2025 Updated Oct 22, 2026
Also listed as: glm-4.5 · glm-4p5 · https://huggingface.co/zai-org/GLM-4.5 · GLM 4.5 · GLM 4.5 (2025-08-22)
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1411 | 148 | Overall, style control on; 23803 votes; 95% CI 1406–1416“glm-4.5” | Official leaderboard | Oct 8, 2026 |
| MMLU-Pro | 84.6% | 36 | Overall accuracy; data source: Self-Reported; 355B | Vendor-reported | Mar 11, 2026 |
| SWE-bench Verified | 64.2% | 42 | agent: Undisclosed“https://huggingface.co/zai-org/GLM-4.5” | Official leaderboard | Jul 28, 2025 |
| SWE-bench Verified | 54.2% | 59 | agent: mini-SWE-agent | Official leaderboard | Aug 22, 2025 |
| Humanity's Last Exam | 8.3% | 35 | Compiled by Epoch AI | Third-party compilation via Epoch AI | Aug 3, 2025 |
FAQ
Where do these GLM-4.5 scores come from?
From Epoch AI Benchmarking Hub, LMArena leaderboard dataset, MMLU-Pro leaderboard (TIGER-Lab), and SWE-bench leaderboards (swebench.com). Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.