Jina OCR vs MinerU
Pick Jina OCR for non-commercial PDF to Markdown work and research, where it scores 83.4 on olmOCR-Bench against 75.2 for MinerU 2.5.4. Pick MinerU when you need a self-hosted tool for mixed PDF and Office files, or when you must run on CPU only. Check the licence first: Jina OCR blocks commercial use without a deal, and MinerU's licence and AGPL-tagged VLM weights also complicate it.
Jina OCRjina/jina-ocr | MinerUopendatalab/mineru | |
|---|---|---|
| Type | Open source | Open source |
| Price | 1 credit / page | 2 credits / page |
| Per 1,000 pages | $1.00 | $2.00 |
| Typical speed | 1.5 s / page | 2.0 s / page |
| Best for | Clean Markdown at low cost | Academic PDFs |
| Inputs | PDF, images | PDF, images |
| Languages | About 100 languages | 80+ languages |
| Max pages per request | 200 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | No | ✓ Yes |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | ✓ Yes | ✓ Yes |
| On OCRRouter | Available | Available |
Prices are what you pay on OCRRouter. Speed is the provider's typical figure, not a benchmark we ran.
Which to choose
Choose Jina OCR if
- It scores 83.4 on olmOCR-Bench, 7.4 points above DeepSeek-OCR, and MinerU 2.5.4 scores 75.2 on the same benchmark.
- Only about 570M of its 3B parameters are active, and it runs at 2.57 pages per second on an A100.
- You can self-host it or call Jina's hosted API, so you can start without running your own GPU.
Choose MinerU if
- It reads PDF, images, DOCX, PPTX, XLSX, EPUB, OFD and HTML in one tool, which suits mixed RAG corpora.
- The basic ONNX tier runs on CPU with about 2 GB RAM, so you need no GPU for batch conversion at basic quality.
- It renders to Markdown, HTML, LaTeX, DOCX, EPUB and JSON, and it suits Chinese and English academic papers.
Key differences
| Jina OCR | MinerU | |
|---|---|---|
| Licence | CC-BY-NC-4.0; commercial use needs a licence from Jina/Elastic | MinerU Open Source License (Apache-2.0 based with extra conditions); VLM weights tagged AGPL-3.0 |
| How you run it | Self-host or Jina's hosted API | Self-host only, as CLI, API or WebUI |
| Hardware | GPU recommended (A100, L4 benchmarks); falls back to CPU in Transformers | Basic tier on CPU with 2 GB RAM; VLM tiers need 16 GB RAM and a GPU with 8 GB+ VRAM |
| Input and output | Page to Markdown, HTML or LaTeX | PDF, images and Office files in; Markdown, HTML, LaTeX, DOCX, EPUB or JSON out |
| olmOCR-Bench overall | 83.4 (authors) | 75.2 for MinerU 2.5.4 (independent) |
| Model size | 3B MoE, about 570M active | MinerU2.5 VLM at 1.2B |
| Known failure | Hosted API returns HTTP 503 on cold start; retry after 30 to 60 seconds. Reported scores use a stricter prompt than the default. | VLM tiers drop merged header rows in tables such as bank statement account lines (issue #5554), and can produce repeated content (issue #5728) |
Sources
- jina-ocr-v1 model card
- Jina model page
- MinerU GitHub README
- olmOCR-Bench table
- Issue #5554 table header rows
- Issue #5728 repetition
What each one is
Jina OCR
Compact page-to-Markdown model from Jina AI, built on the DeepSeek-OCR encoder.
Jina OCR detailsMinerU
Open-source PDF to Markdown pipeline popular for academic and Chinese documents.
MinerU detailsTry both on your own file
Run the file once with each model and compare the output.
Use one, fall back to the other
{"model": "jina/jina-ocr","fallbacks": ["opendatalab/mineru"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Jina OCR alternativesMinerU alternativesBest OCR models in 2026