#2222nd of 42 in Open models
glm-5
Open-weight GLM-5 series LLMs for coding and long-horizon agent tasks
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 7.4k
- License
- Apache-2.0
- Last commit
- Oct 2026
1,972 on 1 Apr 20267,369 on 21 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Release repository for the GLM-5, 5.1, 5.2 and 5.3 model families from Z.ai, with Hugging Face and ModelScope links in FP8 and BF16. The large models are 744B parameters (40B active); GLM-5.3-Flash is 320B (18B active) with hybrid sparse and linear attention. The README covers serving through SGLang, vLLM, Transformers, KTransformers and Unsloth, and fine-tuning with Slime and ms-swift.
Who it is for: Teams with multi-GPU infrastructure serving or fine-tuning large open LLMs
Strengths
- Apache-2.0 repo; weights offered in both FP8 and BF16 for each release
- Serving guides for SGLang, vLLM, Transformers, KTransformers, Unsloth and Ascend NPU
- GLM-5.2 documented with a 1M-token context
- Adjustable reasoning_effort (low, high, max) on GLM-5.3 models
Weaknesses
- 744B-parameter models need multi-GPU servers; README gives no hardware requirements
- Repo holds docs and links only; weights live on Hugging Face and ModelScope
- Benchmark claims are Z.ai's own, including its in-house Code Bench
- reasoning_effort defaults to max; clear_thinking must be set explicitly for chat
What it needs
- GPU required
- Needs SGLang, vLLM, Transformers, KTransformers, Unsloth
- Models: GLM-5, GLM-5.1, GLM-5.2, GLM-5.3, GLM-5.3-Flash
Also in Open models
See all 42| Rank | Project | Score |
|---|---|---|
| 1 | llmfitRust tool that checks which LLMs fit your RAM, CPU and GPU | 87 out of 100 |
| 2 | layaSingle-pass multilingual classifier for typed choice, score and yes/no decisions | 86 out of 100 |
| 3 | timesfmPretrained time-series foundation model for zero-shot forecasting from Google Research | 76 out of 100 |
| #4 | speechPyTorch toolkit for training and deploying ASR, TTS and speech LLM models | 73 out of 100 |
| #5 | yueLyrics-to-song model that writes an editable melody and chord score first | 69 out of 100 |