mineru vs label-studio
Two of the top projects in Data and scraping for AI, side by side: who leads on each score part and who each one suits, then setup, license, activity and what each review found.
mineru
Document parser that converts PDFs, images and Office files to Markdown and JSON
label-studio
Multi-user data labeling tool for audio, text, images, video and time series
Verdict
mineru ranks higher, 80 to 72. Of the 5 score parts we measure for both, mineru leads on 3, label-studio on 1 and 1 is even.
- Adoptionmineru leads by 36 points, 86 to 50: a wide gap.
- FreshnessEven at 100.
- Maintenancemineru leads by 23 points, 98 to 75: a wide gap.
- Easy to runlabel-studio leads by 33 points, 83 to 50: a wide gap.
- Agent-readymineru leads by 70 points, 70 to 0: a wide gap.
Pick mineru if you are
Engineers feeding PDFs and Office documents to LLM pipelines or agents
- Handles scanned PDFs, tables and formulas, with four quality tiers from fast preview to advanced
- Stable doc/tier/page/block locators allow paged reading and citation of long documents
Pick label-studio if you are
ML teams that need to annotate and manage training data
- Handles images, audio, text, HTML, video and time series in one instance
- Installs via pip, Docker, or a Compose stack with Nginx and PostgreSQL
| What we compare | mineru | label-studio |
|---|---|---|
| Score parts, out of 100 | ||
| Adoption | 86, widely used | 50, popular |
| Freshness | 100, active | 100, active |
| Maintenance | 98, healthy | 75, fair |
| Easy to run | 50, easy | 83, very easy |
| Agent-ready | 70, partly | 0, none |
| Facts from GitHub and the README | ||
| Stars | 82.2k | 28.5k |
| License | custom license (read the license) | Apache-2.0 (permissive) |
| Last commit | Oct 2026 | Oct 2026 |
| Last release | Oct 2026 | Sep 2026 |
| Language | Python | TypeScript |
| Docker | Yes | Yes |
| GPU | Optional | Not needed |
| arm64 or Apple Silicon | Mentioned | Mentioned |
mineru
MinerU parses PDFs, scanned images, Office, OpenDocument, EPUB, OFD, HTML and CSV files into Markdown, HTML, LaTeX, DOCX and structured content lists. It offers four parsing tiers (flash, basic, standard, advanced), a local document library with search, and page/block locators for citations. It ships a CLI, Python SDK, API, router and Gradio WebUI, plus an agent skill.
Who it is for: Engineers feeding PDFs and Office documents to LLM pipelines or agents
Strengths
- Handles scanned PDFs, tables and formulas, with four quality tiers from fast preview to advanced
- Stable doc/tier/page/block locators allow paged reading and citation of long documents
- Small models run on ONNX CPU; the VLM can run on llama.cpp, vLLM or LMDeploy
- Documents are not uploaded to the official service unless remote parsing is configured
Weaknesses
- Office, HTML, EPUB, OFD and CSV files only get flash-tier parsing; higher tiers are PDF and image only
- Docker deployment for non-NVIDIA devices is pending an update
- Highest throughput needs an NVIDIA GPU and mineru[full]; Windows needs torch installed separately
- License is reported as NOASSERTION, so terms need checking before use
- GPU optional
- Docker + Compose
- Needs Python >=3.10,<3.15, llama.cpp, vLLM or LMDeploy (VLM runtime), ONNX or Torch
- Models: MinerU VLM, ONNX small models, Torch small models, llama.cpp, vLLM
label-studio
Label Studio is a web-based annotation tool for preparing and correcting ML training data across images, audio, text, HTML, video and time series. Labeling interfaces are defined in a configuration language, with included templates, and results export to various model formats. It can connect to an ML backend for pre-labeling, online learning and active learning, and has a REST API for pipelines.
Who it is for: ML teams that need to annotate and manage training data
Strengths
- Handles images, audio, text, HTML, video and time series in one instance
- Installs via pip, Docker, or a Compose stack with Nginx and PostgreSQL
- ML backend SDK supports pre-labeling, online learning and active learning
- Imports from S3, GCS and files; REST API for pipeline integration
Weaknesses
- Default install stores data in SQLite; PostgreSQL needs the Compose setup
- Some features are tied to paid editions; README links a comparison page
- ML integration needs a separately run backend server
- Windows install requires manually downloading a prebuilt lxml wheel
- no GPU
- Docker + Compose
- Needs PostgreSQL (optional, Compose), Nginx (optional, Compose), MinIO (optional)
- Models: Custom ML backends via Label Studio ML SDK
- port 8080
More in Data and scraping for AI
Building with these?
We help teams pick an open-source AI stack and run it in production. Client work never changes a rank.