Marker vs MonkeyOCR
Pick Marker for self-hosted PDF to Markdown, books and papers with equations, and RAG chunking, or when you want a hosted API option. Pick MonkeyOCR for English and Simplified Chinese PDFs with formulas and tables, in research or evaluation. The main split is licensing and deployment: Marker has a hosted API and free use for startups under $5M, while MonkeyOCR is self-host only and v1 weights need a written license for any commercial use.
Markerdatalab/marker | MonkeyOCRmonkeyocr/monkeyocr | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 14 credits / page | 2 credits / page |
| Per 1,000 pages | $14.00 | $2.00 |
| Typical speed | 2.5 s / page | 1.5 s / page |
| Best for | Books and research papers | Chinese and English PDFs |
| Inputs | PDF, images, DOCX | PDF, images |
| Languages | 90+ languages | Chinese, English |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | ✓ Yes | No |
| 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 Marker if
- It has a hosted API from Datalab with about 5,000 free pages per month, so you can skip running GPUs.
- Fast mode runs on CPU with about 18 MB of ONNX weights at 23.6 pages per second, though olmOCR-bench drops to 66.6 there.
- It outputs Markdown, HTML, JSON block trees, LaTeX and RAG chunks, which suits retrieval pipelines.
Choose MonkeyOCR if
- Pro-3B reports the best overall OmniDocBench edit distance among the models it compared, in English (0.138) and Chinese (0.206).
- It runs on an 8GB RTX 4060 with quantized weights, and pro-1.2B is about 36% faster than pro-3B for about 1.6% lower accuracy.
- It returns bounding boxes and block coordinates in JSON, plus a layout PDF, alongside Markdown.
Key differences
| Marker | MonkeyOCR | |
|---|---|---|
| Licence | Code Apache-2.0; weights under a modified AI Pubs Open Rail-M licence, free for research, personal use and startups under $5M | Code Apache-2.0; v1 weights for research and non-commercial use only, commercial use needs a written licence |
| How you run it | Self-host on GPU, CPU or Apple Silicon, or use Datalab's hosted API | Self-host only, via LMDeploy or Docker |
| Price | Free self-hosted within the weight licence; hosted API has about 5,000 free pages per month, then paid plans | Free for research, self-hosted; commercial use needs a separate licence from the authors |
| Output formats | Markdown, HTML, JSON, chunks, LaTeX, images | Markdown, JSON, LaTeX, HTML, bounding boxes |
| Benchmarks (all self-reported) | olmOCR-bench 83.7 and OmniParseBench 85.9 in vllm mode | OmniDocBench edit distance 0.138 EN / 0.206 ZH and table TEDS 81.5 EN / 87.5 ZH (pro-3B, lower edit distance is better) |
| Speed | 23.6 pages per second in CPU fast mode | Pro-3B about 0.49 pages per second on an RTX 3090 |
| Known limits | Weight licence blocks free commercial use for companies above $5M; CPU fast mode scores 66.6 on olmOCR-bench | Photographed text, handwriting, Traditional Chinese and multilingual text are not fully supported; Docker on 20/30/40-series and V100 needs a patched image |
Sources
What each one is
Marker
Hosted version of the open-source Marker and Surya models. Strong on books and papers.
Marker detailsMonkeyOCR
Open-source document parser that splits work into structure, recognition and relation steps.
MonkeyOCR 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": "datalab/marker","fallbacks": ["monkeyocr/monkeyocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Marker alternativesMonkeyOCR alternativesBest OCR models in 2026