Model · OpenDataLab · Vision
MinerU
An open-source tool that wraps document-parsing models to turn PDFs, scans, Word and PowerPoint files into clean Markdown or JSON, keeping tables, formulas and reading order. Version 4.0 adds a local document library and agent-friendly reading with stable page and block citations.
- Modality
- Vision
- License
- MinerU Open Source License (Open weights)
- Parameter size
- Parsing model MinerU2.5-Pro-2605-1.2B (1.2B), plus small layout and OCR models
- Released
- September 16, 2026
- Last verified
- October 28, 2026
- Runs locally
- Yes
What you need to run it
- Mac
- Yes. Any Apple Silicon Mac (M1 or newer) with at least 8 GB (16 GB recommended) of memory. Newer chips are faster, but memory is what decides whether it runs.
- Windows / Linux PC
- NVIDIA works best. A graphics card with at least 4 GB (8 GB recommended) of video memory.
- No graphics card
- Works. Runs on a normal computer's processor, just slower.
The default install works on Apple Silicon and on a plain CPU; an NVIDIA card is only needed for the fastest vLLM mode.
Numbers are for: basic tier (small models, 8 GB RAM) minimum; standard tier (PyTorch + llama.cpp, 16 GB RAM, 8 GB+ GPU or Apple MPS) recommended. Based on the maker’s own figures; where the maker gives only a graphics-card figure, the Mac memory is our estimate.
Strengths
- Runs on a CPU-only laptop out of the box (ONNX and llama.cpp); an NVIDIA GPU with 8 GB+ VRAM speeds it up
- Keeps tables, formulas and reading order, with nine output formats including Markdown, HTML, LaTeX and DOCX
- Handles PDFs, scanned images, Word and PowerPoint files
- Built for agents and search pipelines: page and block locators let you cite exactly where text came from
- The parsing model itself is Apache 2.0 on Hugging Face
Weaknesses
- Apache 2.0 plus extra terms (MinerU Open Source License): above 100M monthly active users or $20M monthly revenue you need a separate commercial licence
- If you offer an online service built on MinerU, you must say clearly in the product or its public docs that it uses MinerU
- Fast-moving project — twelve 4.0.x patch releases in under a month
Try it
| Where | Type | Notes |
|---|---|---|
| GitHub | local | pip / uv install; CPU or GPU |
| Hugging Face | weights | Parsing model, Apache 2.0 |
Official sources
- GitHub github
- Documentation docs
- MinerU2.5-Pro technical report paper
Change log
- — Added hardware requirements (Mac support, memory, NVIDIA). Source: model card / docs (https://github.com/opendatalab/MinerU (README hardware tier table)).
- — Initial entry. MinerU 4.0.0 released 2026-09-16 (GitHub release); latest is 4.0.12 (2026-10-11). Licence is the MinerU Open Source License — Apache 2.0 with a commercial-licence threshold and an attribution rule for online services. Added 2026-10-28 from the popularity sweep: 82.7k GitHub stars; the parsing model had 204k Hugging Face downloads in the last month.
Esc