Elamur Research⌕
☰
home / papers / 2026 / GLM-OCR Technical Report

GLM-OCR Technical Report

Paper · 2026 · Shuaiqi Duan, Yadong Xue, Weihan Wang et al.

Summary

GLM-OCR addresses a practical gap in document understanding. Large multimodal LLMs read complex documents well but are too heavy and slow for high-concurrency servers or edge devices. Classic OCR pipelines, in contrast, struggle with intricate layouts. Zhipu AI and Tsinghua University present a compact 0.9B-parameter vision-language model. It combines a CogViT image encoder, a light connector and a GLM decoder. OCR output is largely deterministic, so the authors use Multi-Token Prediction (MTP) in both training and decoding. MTP drafts several tokens per step, and its draft layers share parameters to limit extra GPU memory. The system works in two stages: PP-DocLayout-V3 first splits pages into regions, which are then recognised in parallel. Document parsing and key information extraction share one generative model, controlled by prompts. The model ranks first on OmniDocBench v1.5 among the compared systems, including much larger ones. It also scores well on text, formula, table and KIE benchmarks and on in-house production scenarios. Open weights and standard tooling make it a cheap base for domain-specific digitisation.

Key points

  • Architecture: 0.4B CogViT encoder plus 0.5B GLM decoder, built on the GLM-V framework. Layout detection uses PP-DocLayout-V3, and outputs are structured Markdown/JSON.
  • MTP speed-up: the model is trained to predict 10 tokens per step. At inference it averages 5.2 tokens per step, giving roughly 50% higher throughput.
  • Public benchmarks:
    • 94.6 on OmniDocBench v1.5 (rank 1)
    • 94.0 on OCRBench (Text)
    • 96.5 on UniMERNet
    • 85.2 on PubTabNet
    • 86.0 on TEDS_TEST
    • Competitive results on Nanonets-KIE and Handwritten-Forms
  • In-house scenarios (six in total): 91.5 on natural-scene tables, 90.5 on seals and 94.5 on receipt KIE. The multilingual OCR evaluation covers zh, en, fr, es, ru, de, ja and ko; Ukrainian is not listed.
  • Resources:
    • Code and model: zai-org/GLM-OCR on GitHub and Hugging Face
    • Demo: ocr.z.ai
    • Serving: vLLM, SGLang and Ollama
    • Fine-tuning: LLaMA-Factory

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Tsinghua UniversityBeijing · China333 works · 87,019 citations

 Fields, topics and concepts

📚 Shelves 2The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksComputer Vision and Pattern Recognition · 2,869 works

Cite

Cite as Shuaiqi Duan, Yadong Xue, Weihan Wang, Zhe Su, Huan Liu, Sheng Yang, Guobing Gan, Guo Wang, Zihan Wang, Shengdong Yan, Dexin Jin, Yuxuan Zhang, Guohong Wen, Yanfeng Wang, Yutao Zhang, Xiaohan Zhang, Wenyi Hong, Yukuo Cen, Da Yin, Bin Chen, Wenmeng Yu, Xiaotao Gu and Jie Tang. 2026. GLM-OCR Technical Report. arXiv:2603.10910.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2603.10910 · arXiv 2603.10910

@misc{duan2026glm,
  title = {{GLM-OCR Technical Report}},
  author = {Shuaiqi Duan, Yadong Xue, Weihan Wang, Zhe Su, Huan Liu, Sheng Yang, Guobing Gan, Guo Wang, Zihan Wang, Shengdong Yan, Dexin Jin, Yuxuan Zhang, Guohong Wen, Yanfeng Wang, Yutao Zhang, Xiaohan Zhang, Wenyi Hong, Yukuo Cen, Da Yin, Bin Chen, Wenmeng Yu, Xiaotao Gu and Jie Tang},
  year = {2026},
  doi = {10.48550/arxiv.2603.10910},
  eprint = {2603.10910},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2603.10910}
}

References

37 references, 15 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023.
  2. Shuai Bai, Yuxuan Cai, Ruizhe Chen, Keqin Chen, Xionghui Chen, Zesen Cheng, Lianghao Deng, Wei Ding, Chang Gao, Chunjiang Ge, Wenbin Ge, Zhifang Guo, Qidong Huang, Jie Huang, Fei Huang, Binyuan Hui, Shutong Jiang, Zhaohai Li, Mingsheng Li, Mei Li, Kaixin Li, Zicheng Lin, Junyang Lin, Xuejing Liu, Ji… — Qwen3-VL Technical Report
  3. Shuai Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, Sibo Song, Kai Dang, Peng Wang, Shijie Wang, Jun Tang, et al. Qwen2. 5-vl technical report. arXiv preprint arXiv:2502.13923, 2025.
  4. chatdoc com. Ocrflux. Accessed:2025-09-25. https://github.com/chatdoc-com/OCRFlux, 2025. ↗
  5. Cheng Cui, Ting Sun, Suyin Liang, Tingquan Gao, Zelun Zhang, Jiaxuan Liu, Xueqing Wang, Changda Zhou, Hongen Liu, Manhui Lin, Yue Zhang, Yubo Zhang, Handong Zheng, Jing Zhang, Jun Zhang, Yi Liu, Dianhai Yu, and Yanjun Ma. Paddleocr-vl: Boosting multilingual document parsing via a 0.9b ultra-compact… — [arXiv:2510.14528](https://arxiv.org/abs/2510.14528) find
  6. Cheng Cui, Ting Sun, Manhui Lin, Tingquan Gao, Yubo Zhang, Jiaxuan Liu, Xueqing Wang, Zelun Zhang, Changda Zhou, Hongen Liu, Yue Zhang, Wenyu Lv, Kui Huang, Yichao Zhang, Jing Zhang, Jun Zhang, Yi Liu, Dianhai Yu, and Yanjun Ma. Paddleocr 3.0 technical report, 2025. URL https://arxiv.org/abs/2507.05… — [arXiv:2507.05595](https://arxiv.org/abs/2507.05595) ↗
  7. Cheng Cui, Ting Sun, Suyin Liang, Tingquan Gao, Zelun Zhang, Jiaxuan Liu, Xueqing Wang, Changda Zhou, Hongen Liu, Manhui Lin, Yue Zhang, Yubo Zhang, Yi Liu, Dianhai Yu, and Yanjun Ma. Paddleocr-vl-1.5: Towards a multi-task 0.9b vlm for robust in-the-wild document parsing, 2026. URL https://arxiv.org… — [arXiv:2601.21957](https://arxiv.org/abs/2601.21957) ↗
  8. Hao Feng, Shu Wei, Xiang Fei, Wei Shi, Yingdong Han, Lei Liao, Jinghui Lu, Binghong Wu, Qi Liu, Chunhui Lin, et al. Dolphin: Document image parsing via heterogeneous anchor prompting. arXiv preprint arXiv:2505.14059, 2025. find
  9. Team GLM, Aohan Zeng, Bin Xu, Bowen Wang, Chenhui Zhang, Da Yin, Diego Rojas, Guanyu Feng, Hanlin Zhao, Hanyu Lai, Hao Yu, Hongning Wang, Jiadai Sun, Jiajie Zhang, Jiale Cheng, Jiayi Gui, Jie Tang, Jing Zhang, Juanzi Li, Lei Zhao, Lindong Wu, Lucen Zhong, Mingdao Liu, Minlie Huang, Peng Zhang, Qinka… — ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools
  10. Google DeepMind. Gemini 2.5. https://blog.google/technology/ google-deepmind/gemini-model-thinking-updates-march-2025/, 2025. ↗
  11. Google DeepMind. Gemini 3.0. https://blog.google/ products-and-platforms/products/gemini/gemini-3-collection/, 2025. ↗
  12. JaidedAI. Easyocr. https://github.com/JaidedAI/EasyOCR, 2020. ↗
  13. Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph E. Gonzalez, Hao Zhang, and Ion Stoica. Efficient memory management for large language model serving with pagedattention. In Proceedings of the ACM SIGOPS 29th Symposium on Operating Systems Principles, 2023.
  14. Zhang Li, Yuliang Liu, Qiang Liu, Zhiyin Ma, Ziyang Zhang, Shuo Zhang, Zidun Guo, Jiarui Zhang, Xinyu Wang, and Xiang Bai. Monkeyocr: Document parsing with a structure-recognition-relation triplet paradigm. arXiv preprint arXiv:2506.05218, 2025. find
  15. Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, et al. Deepseek-v3 technical report. arXiv preprint arXiv:2412.19437, 2024. find
  16. Yuan Liu, Zhongyin Zhao, Le Tian, Haicheng Wang, Xubing Ye, Yangxiu You, Zilin Yu, Chuhan Wu, Xiao Zhou, Yang Yu, et al. Points-reader: Distillation-free adaptation of vision-language models for document conversion. arXiv preprint arXiv:2509.01215, 2025. find
  17. Yuliang Liu, Zhang Li, Mingxin Huang, Biao Yang, Wenwen Yu, Chunyuan Li, Xu-Cheng Yin, Cheng-Lin Liu, Lianwen Jin, and Xiang Bai. Ocrbench: on the hidden mystery of ocr in large multimodal models. Science China Information Sciences, 67(12), December 2024. ISSN 1869-1919. doi: 10.1007/s11432-024-4235…
  18. Souvik Mandal, Nayancy Gupta, Ashish Talewar, Paras Ahuja, Prathamesh Juvatkar, and Gourinath Banda. Idpleaderboard: A unified leaderboard for intelligent document processing tasks. https://idp-leaderboard.org, 2025. ↗
  19. Souvik Mandal, Ashish Talewar, Paras Ahuja, and Prathamesh Juvatkar. Nanonets-ocr-s: A model for transforming documents into structured markdown with intelligent content recognition and semantic tagging, 2025. find
  20. Mistral AI Team. Mistral-ocr. https://mistral.ai/news/mistral-ocr?utm_ source=ai-bot.cn, 2025. ↗
  21. Junbo Niu, Zheng Liu, Zhuangcheng Gu, Bin Wang, Linke Ouyang, Zhiyuan Zhao, Tao Chu, Tianyao He, Fan Wu, Qintong Zhang, et al. Mineru2. 5: A decoupled vision-language model for efficient high-resolution document parsing. arXiv preprint arXiv:2509.22186, 2025.
  22. OpenAI. Gpt-5.2 system card, 2025. URL https://cdn.openai.com/pdf/ 3a4153c8-c748-4b71-8e31-aecbde944f8d/oai_5_2_system-card.pdf. ↗
  23. opendatalab. Mineru2.0-2505-0.9b. MinerU2.0-2505-0.9B, 2025. https://huggingface.co/opendatalab/ ↗
  24. Linke Ouyang, Yuan Qu, Hongbin Zhou, Jiawei Zhu, Rui Zhang, Qunshu Lin, Bin Wang, Zhiyuan Zhao, Man Jiang, Xiaomeng Zhao, Jin Shi, Fan Wu, Pei Chu, Minghao Liu, Zhenxiang Li, Chao Xu, Bo Zhang, Botian Shi, Zhongying Tu, and Conghui He. Omnidocbench: Benchmarking diverse pdf document parsing with com… — [arXiv:2412.07626](https://arxiv.org/abs/2412.07626) find
  25. Vik Paruchuri. Marker. https://github.com/datalab-to/marker, 2025. Accessed: 2025-09-25. ↗
  26. Jake Poznanski, Jon Borchardt, Jason Dunkelberger, Regan Huff, Daniel Lin, Aman Rangapur, Christopher Wilhelm, Kyle Lo, and Luca Soldaini. olmocr: Unlocking trillions of tokens in pdfs with vision language models. arXiv preprint arXiv:2502.18443, 2025.
  27. rednote-hilab. dots.ocr: Multilingual document layout parsing in a single vision-language model, 2025. find
  28. Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, YK Li, Y Wu, et al. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. arXiv preprint arXiv:2402.03300, 2024.
  29. Ray Smith. An overview of the tesseract ocr engine. In Ninth international conference on document analysis and recognition (ICDAR 2007), volume 2, pp. 629–633. IEEE, 2007. find
  30. ByteDance Seed Team. Seed1.5-vl technical report. arXiv preprint arXiv:2505.07062, 2025.
  31. V Team, Wenyi Hong, Wenmeng Yu, Xiaotao Gu, Guo Wang, Guobing Gan, Haomiao Tang, Jiale Cheng, Ji Qi, Junhui Ji, Lihang Pan, Shuaiqi Duan, Weihan Wang, Yan Wang, Yean Cheng, Zehai He, Zhe Su, Zhen Yang, Ziyang Pan, Aohan Zeng, Baoxu Wang, Bin Chen, Boyan Shi, Changyu Pang, Chenhui Zhang, Da Yin, Fan… — GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
  32. Bin Wang, Zhuangcheng Gu, Guang Liang, Chao Xu, Bo Zhang, Botian Shi, and Conghui He. Unimernet: A universal network for real-world mathematical expression recognition, 2024. URL https://arxiv.org/abs/2404.15254. ↗
  33. Weiyun Wang, Zhangwei Gao, Lixin Gu, Hengjun Pu, Long Cui, Xingguang Wei, Zhaoyang Liu, Linglin Jing, Shenglong Ye, Jie Shao, et al. Internvl3. 5: Advancing open-source multimodal models in versatility, reasoning, and efficiency. arXiv preprint arXiv:2508.18265, 2025.
  34. Haoran Wei, Yaofeng Sun, and Yukun Li. Deepseek-ocr: Contexts optical compression. arXiv preprint arXiv:2510.18234, 2025.
  35. Aohan Zeng, Xin Lv, Zhenyu Hou, Zhengxiao Du, Qinkai Zheng, Bin Chen, Da Yin, Chendi Ge, Chengxing Xie, Cunxiang Wang, et al. Glm-5: from vibe coding to agentic engineering. arXiv preprint arXiv:2602.15763, 2026.
  36. Xu Zhong, Elaheh ShafieiBavani, and Antonio Jimeno Yepes. Image-based table recognition: data, model, and evaluation. arXiv preprint arXiv:1911.10683, 2019. — [arXiv:1911.10683](https://arxiv.org/abs/1911.10683) find
  37. Jinguo Zhu, Weiyun Wang, Zhe Chen, Zhaoyang Liu, Shenglong Ye, Lixin Gu, Hao Tian, Yuchen Duan, Weijie Su, Jie Shao, et al. Internvl3: Exploring advanced training and test-time recipes for open-source multimodal models. arXiv preprint arXiv:2504.10479, 2025.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/duan-2026-glm-ocr-technical-report/