Model Intelligence Sheet
mungert/mineru2.5-2509-1.2b-gguf overview
Comprehensive model page for mungert/mineru2.5-2509-1.2b-gguf
Downloads
2,184
Likes
0
Pipeline
image-text-to-text
Library
transformers
Visibility
Public
Access
Open
Repository Files & Downloads
19 files detected
Direct downloads for all repository files
| File | Type | Quantization | Size | Link |
|---|---|---|---|---|
| MinerU2.5-2509-1.2B-bf16.gguf | GGUF | BF16 | 948.10 MB | Download |
| MinerU2.5-2509-1.2B-f16_q8_0.gguf | GGUF | F16 | 741.38 MB | Download |
| MinerU2.5-2509-1.2B-imatrix.gguf | GGUF | — | 0.96 MB | Download |
| MinerU2.5-2509-1.2B-iq3_m.gguf | GGUF | IQ3_M | 295.56 MB | Download |
| MinerU2.5-2509-1.2B-iq3_xs.gguf | GGUF | IQ3_XS | 284.88 MB | Download |
| MinerU2.5-2509-1.2B-iq3_xxs.gguf | GGUF | IQ3_XXS | 284.88 MB | Download |
| MinerU2.5-2509-1.2B-iq4_nl.gguf | GGUF | IQ4_NL | 271.42 MB | Download |
| MinerU2.5-2509-1.2B-iq4_xs.gguf | GGUF | IQ4_XS | 292.64 MB | Download |
| MinerU2.5-2509-1.2B-q3_k_m.gguf | GGUF | Q3_K_M | 297.91 MB | Download |
| MinerU2.5-2509-1.2B-q3_k_s.gguf | GGUF | Q3_K_S | 295.27 MB | Download |
| MinerU2.5-2509-1.2B-q4_0.gguf | GGUF | — | 335.84 MB | Download |
| MinerU2.5-2509-1.2B-q4_1.gguf | GGUF | — | 308.48 MB | Download |
| MinerU2.5-2509-1.2B-q4_k_m.gguf | GGUF | Q4_K_M | 376.07 MB | Download |
| MinerU2.5-2509-1.2B-q4_k_s.gguf | GGUF | Q4_K_S | 328.60 MB | Download |
| MinerU2.5-2509-1.2B-q5_0.gguf | GGUF | — | 378.50 MB | Download |
| MinerU2.5-2509-1.2B-q5_1.gguf | GGUF | — | 399.82 MB | Download |
| MinerU2.5-2509-1.2B-q5_k_m.gguf | GGUF | Q5_K_M | 400.62 MB | Download |
| MinerU2.5-2509-1.2B-q6_k_m.gguf | GGUF | Q6_K_M | 482.31 MB | Download |
| MinerU2.5-2509-1.2B-q8_0.gguf | GGUF | — | 506.46 MB | Download |
Model Details Live
Metadata Inspector
Normalized metadata (stored in metadata_json)
{
"metadata": {},
"card_data": {
"license": "agpl-3.0",
"language": [
"zh",
"en"
],
"pipeline_tag": "image-text-to-text",
"library_name": "transformers",
"frontmatter": {
"license": "agpl-3.0",
"language": [
"zh",
"en"
],
"pipeline_tag": "image-text-to-text",
"library_name": "transformers"
},
"hero_image_url": "https://raw.githubusercontent.com/opendatalab/MinerU/master/docs/images/MinerU-logo.png",
"summary": "",
"quick_links": [],
"benchmark_table_html": "",
"readme_markdown": "---\nlicense: agpl-3.0\nlanguage:\n- zh\n- en\npipeline_tag: image-text-to-text\nlibrary_name: transformers\n---\n\n# <span style=\"color: #7FFF7F;\">MinerU2.5-2509-1.2B GGUF Models</span>\n\n\n## <span style=\"color: #7F7FFF;\">Model Generation Details</span>\n\nThis model was generated using [llama.cpp](https://github.com/ggerganov/llama.cpp) at commit [`ee09828cb`](https://github.com/ggerganov/llama.cpp/commit/ee09828cb057460b369576410601a3a09279e23c).\n\n\n\n\n\n\n---\n\n<a href=\"https://readyforquantum.com/huggingface_gguf_selection_guide.html\" style=\"color: #7FFF7F;\">\n Click here to get info on choosing the right GGUF model format\n</a>\n\n---\n\n\n\n<!--Begin Original Model Card-->\n\n\n\n<div align=\"center\">\n\n<p align=\"center\">\n <img src=\"https://raw.githubusercontent.com/opendatalab/MinerU/master/docs/images/MinerU-logo.png\" width=\"300\"/>\n<p>\n\n<h1 align=\"center\" style=\"font-size: 28px\">\nMinerU2.5: A Decoupled Vision-Language Model for Efficient High-Resolution Document Parsing\n</h1>\n\n[](https://github.com/opendatalab/MinerU/)\n[](https://huggingface.co/opendatalab/MinerU2.5-2509-1.2B)\n[](https://modelscope.cn/models/OpenDataLab/MinerU2.5-2509-1.2B)\n[](https://huggingface.co/spaces/opendatalab/MinerU)\n[](https://www.modelscope.cn/studios/OpenDataLab/MinerU)\n\n\n<div align=\"center\">\n <a href=\"https://mineru.net/OpenSourceTools/Extractor\" target=\"_blank\" rel=\"noopener noreferrer\"><strong>🚀 Official Demo</strong></a> | \n <a href=\"https://arxiv.org/abs/2509.22186\" target=\"_blank\" rel=\"noopener noreferrer\"><strong>📄 Technical Report</strong></a> \n</div>\n\n</div>\n\n---\n\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/performance.jpeg\"/>\n<p>\n\n# Introduction\n<!-- We present **MinerU2.5**, a 1.2B-parameter VLM-based document parsing model that delivers state-of-the-art accuracy with high efficiency. It adopts a coarse-to-fine, two-stage parsing strategy. A large-scale, diverse data engine supports both pretraining and fine-tuning, enabling robust performance across document types. -->\n**MinerU2.5** is a 1.2B-parameter vision-language model for document parsing that achieves state-of-the-art accuracy with high computational efficiency. It adopts a two-stage parsing strategy: first conducting efficient global layout analysis on downsampled images, then performing fine-grained content recognition on native-resolution crops for text, formulas, and tables. Supported by a large-scale, diverse data engine for pretraining and fine-tuning, MinerU2.5 consistently outperforms both general-purpose and domain-specific models across multiple benchmarks while maintaining low computational overhead.\n\n## Key Improvements\n<!-- - **More Precise Layout Detection:** Faithfully preserves non-body elements such as headers, footers, and page numbers, ensuring comprehensive content integrity.\n- **Significantly Improved Body Text Recognition:** Produces more standardized text formatting with clearly discernible structures for lists, references, and other elements.\n- **Breakthroughs in Formula Parsing:** Delivers high-quality parsing of complex, lengthy mathematical formulae and accurately recognizes mixed-language (Chinese-English) equations.\n- **Enhanced Robustness in Table Parsing:** Effortlessly handles challenging cases, including rotated tables, borderless tables, and tables with partial borders. -->\n\n- **Comprehensive and Granular Layout Analysis:** It not only preserves non-body elements like headers, footers, and page numbers to ensure full content integrity, but also employs a refined and standardized labeling schema. This enables a clearer, more structured representation of elements such as lists, references, and code blocks.\n- **Breakthroughs in Formula Parsing:** Delivers high-quality parsing of complex, lengthy mathematical formulae and accurately recognizes mixed-language (Chinese-English) equations.\n- **Enhanced Robustness in Table Parsing:** Effortlessly handles challenging cases, including rotated tables, borderless tables, and tables with partial borders.\n\n\n# Quick Start\nFor convenience, we provide `mineru-vl-utils`, a Python package that simplifies the process of sending requests and handling responses from MinerU2.5 Vision-Language Model. Here we give some examples to use MinerU2.5. For more information and usages, please refer to [mineru-vl-utils](https://github.com/opendatalab/mineru-vl-utils/tree/main).\n\n📌 We strongly recommend using vllm for inference, as the `vllm-async-engine` can achieve a concurrent inference speed of **2.12 fps** on one A100.\n\n## Install packages\n```bash\n# For `transformers` backend\npip install \"mineru-vl-utils[transformers]\"\n# For `vllm-engine` and `vllm-async-engine` backend\npip install \"mineru-vl-utils[vllm]\"\n```\n\n## `transformers` Example\n\n```python\nfrom transformers import AutoProcessor, Qwen2VLForConditionalGeneration\nfrom PIL import Image\nfrom mineru_vl_utils import MinerUClient\n\n# for transformers>=4.56.0\nmodel = Qwen2VLForConditionalGeneration.from_pretrained(\n \"opendatalab/MinerU2.5-2509-1.2B\",\n dtype=\"auto\", # use `torch_dtype` instead of `dtype` for transformers<4.56.0\n device_map=\"auto\"\n)\n\nprocessor = AutoProcessor.from_pretrained(\n \"opendatalab/MinerU2.5-2509-1.2B\",\n use_fast=True\n)\n\nclient = MinerUClient(\n backend=\"transformers\",\n model=model,\n processor=processor\n)\n\nimage = Image.open(\"/path/to/the/test/image.png\")\nextracted_blocks = client.two_step_extract(image)\nprint(extracted_blocks)\n```\n\n## `vllm-engine` Example (Recommended!)\n\n```python\nfrom vllm import LLM\nfrom PIL import Image\nfrom mineru_vl_utils import MinerUClient\nfrom mineru_vl_utils import MinerULogitsProcessor # if vllm>=0.10.1\n\nllm = LLM(\n model=\"opendatalab/MinerU2.5-2509-1.2B\",\n logits_processors=[MinerULogitsProcessor] # if vllm>=0.10.1\n)\n\nclient = MinerUClient(\n backend=\"vllm-engine\",\n vllm_llm=llm\n)\n\nimage = Image.open(\"/path/to/the/test/image.png\")\nextracted_blocks = client.two_step_extract(image)\nprint(extracted_blocks)\n```\n\n## `vllm-async-engine` Example (Recommended!)\n\n```python\nimport io\nimport asyncio\nimport aiofiles\n\nfrom vllm.v1.engine.async_llm import AsyncLLM\nfrom vllm.engine.arg_utils import AsyncEngineArgs\nfrom PIL import Image\nfrom mineru_vl_utils import MinerUClient\nfrom mineru_vl_utils import MinerULogitsProcessor # if vllm>=0.10.1\n\nasync_llm = AsyncLLM.from_engine_args(\n AsyncEngineArgs(\n model=\"opendatalab/MinerU2.5-2509-1.2B\",\n logits_processors=[MinerULogitsProcessor] # if vllm>=0.10.1\n )\n)\n\nclient = MinerUClient(\n backend=\"vllm-async-engine\",\n vllm_async_llm=async_llm,\n)\n\nasync def main():\n image_path = \"/path/to/the/test/image.png\"\n async with aiofiles.open(image_path, \"rb\") as f:\n image_data = await f.read()\n image = Image.open(io.BytesIO(image_data))\n extracted_blocks = await client.aio_two_step_extract(image)\n print(extracted_blocks)\n\nasyncio.run(main())\n\nasync_llm.shutdown()\n```\n\n# Model Architecture\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/Mineru25_framework.jpeg\"/>\n<p>\n\n# Performance on OmniDocBench\n## Across Different Elements\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/omnidocbench_element.jpeg\"/>\n<p>\n\n## Across Various Document Types\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/omnidocbench_type.jpeg\"/>\n<p>\n\n\n# Case Demonstration\n## Full-Document Parsing across Various Doc-Types\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/PDF-Type-1_page_1.png\"/>\n<p>\n\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/PDF-Type-2_page_1.png\"/>\n<p>\n\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5//PDF-Type-3_page_1.png\"/>\n<p>\n\n## Table Recognition\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/Table-Module-1_page_1.png\"/>\n<p>\n\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/Table-Module-2_page_1.png\"/>\n<p>\n\n## Formula Recognition\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/Formula-Module-1_page_1.png\"/>\n<p>\n\n<p align=\"center\">\n <img alt=\"Image\" src=\"https://hotelll.github.io/MinerU2.5/Formula-Module-2_page_1.png\"/>\n<p>\n\n\n# Acknowledgements\nWe would like to thank [Qwen Team](https://github.com/QwenLM), [vLLM](https://github.com/vllm-project/vllm), [OmniDocBench](https://github.com/opendatalab/OmniDocBench), [UniMERNet](https://github.com/opendatalab/UniMERNet), [PaddleOCR](https://github.com/PaddlePaddle/PaddleOCR), [DocLayout-YOLO](https://github.com/opendatalab/DocLayout-YOLO) for providing valuable code and models. We also appreciate everyone's contribution to this open-source project!\n\n\n# Citation\n\nIf you find our work useful in your research, please consider giving a star ⭐ and citation 📝 :\n\n```BibTeX\n@misc{niu2025mineru25decoupledvisionlanguagemodel,\n title={MinerU2.5: A Decoupled Vision-Language Model for Efficient High-Resolution Document Parsing}, \n author={Junbo Niu and Zheng Liu and Zhuangcheng Gu and Bin Wang and Linke Ouyang and Zhiyuan Zhao and Tao Chu and Tianyao He and Fan Wu and Qintong Zhang and Zhenjiang Jin and others},\n year={2025},\n eprint={2509.22186},\n archivePrefix={arXiv},\n primaryClass={cs.CV},\n url={https://arxiv.org/abs/2509.22186}, \n}\n```\n\n<!--End Original Model Card-->\n\n---\n\n# <span id=\"testllm\" style=\"color: #7F7FFF;\">🚀 If you find these models useful</span>\n\nHelp me test my **AI-Powered Quantum Network Monitor Assistant** with **quantum-ready security checks**: \n\n👉 [Quantum Network Monitor](https://readyforquantum.com/?assistant=open&utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) \n\n\nThe full Open Source Code for the Quantum Network Monitor Service available at my github repos ( repos with NetworkMonitor in the name) : [Source Code Quantum Network Monitor](https://github.com/Mungert69). You will also find the code I use to quantize the models if you want to do it yourself [GGUFModelBuilder](https://github.com/Mungert69/GGUFModelBuilder)\n\n💬 **How to test**: \n Choose an **AI assistant type**: \n - `TurboLLM` (GPT-4.1-mini) \n - `HugLLM` (Hugginface Open-source models) \n - `TestLLM` (Experimental CPU-only) \n\n### **What I’m Testing** \nI’m pushing the limits of **small open-source models for AI network monitoring**, specifically: \n- **Function calling** against live network services \n- **How small can a model go** while still handling: \n - Automated **Nmap security scans** \n - **Quantum-readiness checks** \n - **Network Monitoring tasks** \n\n🟡 **TestLLM** – Current experimental model (llama.cpp on 2 CPU threads on huggingface docker space): \n- ✅ **Zero-configuration setup** \n- ⏳ 30s load time (slow inference but **no API costs**) . No token limited as the cost is low.\n- 🔧 **Help wanted!** If you’re into **edge-device AI**, let’s collaborate! \n\n### **Other Assistants** \n🟢 **TurboLLM** – Uses **gpt-4.1-mini** :\n- **It performs very well but unfortunatly OpenAI charges per token. For this reason tokens usage is limited. \n- **Create custom cmd processors to run .net code on Quantum Network Monitor Agents**\n- **Real-time network diagnostics and monitoring**\n- **Security Audits**\n- **Penetration testing** (Nmap/Metasploit) \n\n🔵 **HugLLM** – Latest Open-source models: \n- 🌐 Runs on Hugging Face Inference API. Performs pretty well using the lastest models hosted on Novita.\n\n### 💡 **Example commands you could test**: \n1. `\"Give me info on my websites SSL certificate\"` \n2. `\"Check if my server is using quantum safe encyption for communication\"` \n3. `\"Run a comprehensive security audit on my server\"`\n4. '\"Create a cmd processor to .. (what ever you want)\" Note you need to install a [Quantum Network Monitor Agent](https://readyforquantum.com/Download/?utm_source=huggingface&utm_medium=referral&utm_campaign=huggingface_repo_readme) to run the .net code on. This is a very flexible and powerful feature. Use with caution!\n\n### Final Word\n\nI fund the servers used to create these model files, run the Quantum Network Monitor service, and pay for inference from Novita and OpenAI—all out of my own pocket. All the code behind the model creation and the Quantum Network Monitor project is [open source](https://github.com/Mungert69). Feel free to use whatever you find helpful.\n\nIf you appreciate the work, please consider [buying me a coffee](https://www.buymeacoffee.com/mahadeva) ☕. Your support helps cover service costs and allows me to raise token limits for everyone.\n\nI'm also open to job opportunities or sponsorship.\n\nThank you! 😊\n",
"related_quantizations": []
},
"tags": [
"transformers",
"gguf",
"image-text-to-text",
"zh",
"en",
"arxiv:2509.22186",
"license:agpl-3.0",
"endpoints_compatible",
"region:us",
"conversational"
],
"likes": 0,
"downloads": 2184,
"gated": false,
"private": false,
"last_modified": "2025-10-20T04:41:28.000Z",
"created_at": "2025-10-20T04:06:21.000Z",
"pipeline_tag": "image-text-to-text",
"library_name": "transformers"
}
Source payload excerpt (from Hugging Face API)
{
"_id": "68f5b53dc6651977ecb0823a",
"id": "Mungert/MinerU2.5-2509-1.2B-GGUF",
"modelId": "Mungert/MinerU2.5-2509-1.2B-GGUF",
"sha": "9886338941b18153b5c5f22809edce055206b440",
"createdAt": "2025-10-20T04:06:21.000Z",
"lastModified": "2025-10-20T04:41:28.000Z",
"author": "Mungert",
"downloads": 2184,
"likes": 0,
"gated": false,
"private": false,
"pipeline_tag": "image-text-to-text",
"library_name": "transformers",
"siblings_count": 25
}