GLM-5.1
zhipuai-glm-5-1
fields
| nameREQ | GLM-5.1 | www.buildfastwithai.com | 2026-05-03 |
| vendorREQ | Z.ai / Zhipu AI | www.buildfastwithai.com | 2026-05-03 |
| released_at | 2026-04-07 | www.buildfastwithai.com | 2026-05-03 |
| modality | text, code | www.buildfastwithai.com | 2026-05-03 |
| access | open-weights | www.buildfastwithai.com | 2026-05-03 |
| context_window | |||
| parameters | 744B total / 40B active | www.buildfastwithai.com | 2026-05-03 |
| benchmark_score | 58.4 | www.buildfastwithai.com | 2026-05-03 |
| benchmark_name | SWE-Bench Pro | www.buildfastwithai.com | 2026-05-03 |
| announcement_url | https://huggingface.co/THUDM/GLM-5.1 | www.buildfastwithai.com | 2026-05-03 |
history · 9 fields · 9 revisions
name1 revision
GLM-5.1
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
vendor1 revision
Z.ai / Zhipu AI
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
released_at1 revision
2026-04-07
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
modality1 revision
text, code
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
access1 revision
open-weights
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
parameters1 revision
744B total / 40B active
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
benchmark_score1 revision
58.4
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
benchmark_name1 revision
SWE-Bench Pro
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.
announcement_url1 revision
https://huggingface.co/THUDM/GLM-5.1
current
On April 7, 2026, GLM-5.1 became the first open-weight model in history to top the SWE-bench Pro leaderboard, scoring 58.4%. GLM-5.1 is a 744 billion parameter Mixture-of-Experts model with 40 billion active parameters. Trained entirely on 100,000 Huawei Ascend 910B chips.