Mudora

Mudora

<p align="center"><em>The book that reads the old script.</em></p>

Mudora is GLM-5.2 under a different name. Same weights, byte for byte. It runs as the main brain inside LEGION, the Destin L Mincy software & AI agency multi-agent orchestration system, and this repo is its permanent address.

The name comes from the Book of Mudora in The Legend of Zelda: A Link to the Past, the tome that turns ancient Hylian script into something a person can read. That is the whole job of a coding model: take intent written in one language, return working text in another. The name fit too well to pass up.

Why this repo exists

Upstream slugs change. Free tiers vanish. Endpoints get deprecated without notice. When a pipeline hardcodes someone else's model string, every one of those events is an outage.

LEGION references mudora. What mudora points at gets decided here, once, and everything downstream keeps working. That is the entire play: naming stability as an infrastructure layer, not a branding exercise.

Second reason: if fine-tuned variants ever ship, this is where they live. Until this card says otherwise, assume the weights here match upstream exactly. That sentence is the tripwire. If it disappears, something changed.

The model underneath

GLM-5.2 is Zhipu AI's flagship for long-horizon coding and agentic work: a sparse Mixture-of-Experts model, roughly 744B total parameters, about 40B active per token. What matters in practice:

• 1M-token context that holds up across long-horizon tasks instead of degrading past the marketing number
• Coding with selectable effort levels, so you trade quality against latency deliberately instead of getting one fixed behavior
• IndexShare sparse attention: one indexer reused across every four sparse attention layers, cutting per-token FLOPs by 2.9x at 1M context. The improved MTP layer raises speculative decoding acceptance length by up to 20%
• MIT license. No regional carve-outs, no usage tiers, no permission slips

Full details: GLM-5.2 blog | GLM-5 technical report | GitHub

Benchmarks

These are Zhipu's published numbers, reproduced verbatim. Mudora inherits them because the weights are identical. Nobody has independently evaluated this mirror, and none of these numbers are ours. Harness settings live in the upstream model card and technical report.

Results marked with * are from the full HLE set rather than the text-only subset.

Serving Mudora

The weights are GLM-5.2 weights, so every upstream deployment path works as documented:

• SGLang (v0.5.13.post1+): cookbook
• vLLM (v0.23.0+): recipes
• Transformers (v0.5.12+): docs
• KTransformers (v0.5.12+): tutorial
• Unsloth (v0.1.47-beta+): guide
• Ascend NPU platforms (vLLM-Ascend, xLLM, SGLang): notes

Before you click download: MoE saves you compute, not memory. Only 40B parameters fire per token, but all 744B have to live somewhere, and that somewhere is hundreds of gigabytes of fast memory. Self-hosting this model is a server-class commitment. On consumer hardware, point at a hosted endpoint and keep mudora as the alias. Build accordingly.

Attribution and license

All credit for this model belongs to the GLM-5 team at Zhipu AI and Tsinghua University. This repo exists for naming stability, not to claim their work. The MIT license and original copyright notice are retained in this repo's LICENSE file, as the license requires.

If you use this model in research, cite the original technical report:

@misc{glm5team2026glm5vibecodingagentic,
      title={GLM-5: from Vibe Coding to Agentic Engineering},
      author={GLM-5-Team and : and Aohan Zeng and Xin Lv and Zhenyu Hou and Zhengxiao Du and Qinkai Zheng and Bin Chen and Da Yin and Chendi Ge and Chenghua Huang and Chengxing Xie and Chenzheng Zhu and Congfeng Yin and Cunxiang Wang and Gengzheng Pan and Hao Zeng and Haoke Zhang and Haoran Wang and Huilong Chen and Jiajie Zhang and Jian Jiao and Jiaqi Guo and Jingsen Wang and Jingzhao Du and Jinzhu Wu and Kedong Wang and Lei Li and Lin Fan and Lucen Zhong and Mingdao Liu and Mingming Zhao and Pengfan Du and Qian Dong and Rui Lu and Shuang-Li and Shulin Cao and Song Liu and Ting Jiang and Xiaodong Chen and Xiaohan Zhang and Xuancheng Huang and Xuezhen Dong and Yabo Xu and Yao Wei and Yifan An and Yilin Niu and Yitong Zhu and Yuanhao Wen and Yukuo Cen and Yushi Bai and Zhongpei Qiao and Zihan Wang and Zikang Wang and Zilin Zhu and Ziqiang Liu and Zixuan Li and Bojie Wang and Bosi Wen and Can Huang and Changpeng Cai and Chao Yu and Chen Li and Chengwei Hu and Chenhui Zhang and Dan Zhang and Daoyan Lin and Dayong Yang and Di Wang and Ding Ai and Erle Zhu and Fangzhou Yi and Feiyu Chen and Guohong Wen and Hailong Sun and Haisha Zhao and Haiyi Hu and Hanchen Zhang and Hanrui Liu and Hanyu Zhang and Hao Peng and Hao Tai and Haobo Zhang and He Liu and Hongwei Wang and Hongxi Yan and Hongyu Ge and Huan Liu and Huanpeng Chu and Jia'ni Zhao and Jiachen Wang and Jiajing Zhao and Jiamin Ren and Jiapeng Wang and Jiaxin Zhang and Jiayi Gui and Jiayue Zhao and Jijie Li and Jing An and Jing Li and Jingwei Yuan and Jinhua Du and Jinxin Liu and Junkai Zhi and Junwen Duan and Kaiyue Zhou and Kangjian Wei and Ke Wang and Keyun Luo and Laiqiang Zhang and Leigang Sha and Liang Xu and Lindong Wu and Lintao Ding and Lu Chen and Minghao Li and Nianyi Lin and Pan Ta and Qiang Zou and Rongjun Song and Ruiqi Yang and Shangqing Tu and Shangtong Yang and Shaoxiang Wu and Shengyan Zhang and Shijie Li and Shuang Li and Shuyi Fan and Wei Qin and Wei Tian and Weining Zhang and Wenbo Yu and Wenjie Liang and Xiang Kuang and Xiangmeng Cheng and Xiangyang Li and Xiaoquan Yan and Xiaowei Hu and Xiaoying Ling and Xing Fan and Xingye Xia and Xinyuan Zhang and Xinze Zhang and Xirui Pan and Xu Zou and Xunkai Zhang and Yadi Liu and Yandong Wu and Yanfu Li and Yidong Wang and Yifan Zhu and Yijun Tan and Yilin Zhou and Yiming Pan and Ying Zhang and Yinpei Su and Yipeng Geng and Yong Yan and Yonglin Tan and Yuean Bi and Yuhan Shen and Yuhao Yang and Yujiang Li and Yunan Liu and Yunqing Wang and Yuntao Li and Yurong Wu and Yutao Zhang and Yuxi Duan and Yuxuan Zhang and Zezhen Liu and Zhengtao Jiang and Zhenhe Yan and Zheyu Zhang and Zhixiang Wei and Zhuo Chen and Zhuoer Feng and Zijun Yao and Ziwei Chai and Ziyuan Wang and Zuzhou Zhang and Bin Xu and Minlie Huang and Hongning Wang and Juanzi Li and Yuxiao Dong and Jie Tang},
      year={2026},
      eprint={2602.15763},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2602.15763},
}