MonkeyOCR vs Chunkr
Pick MonkeyOCR for self-hosted parsing of English and Simplified Chinese PDFs, such as academic papers with formulas and tables, in research or evaluation. Pick Chunkr for RAG chunking of PDFs, slides, Word files and images, or when you want a hosted API. The main split is licensing and scope: MonkeyOCR v1 weights are non-commercial without a written license, while Chunkr is AGPL-3.0 or commercial and has a cloud API.
MonkeyOCRmonkeyocr/monkeyocr | Chunkrchunkr/chunkr | |
|---|---|---|
| Type | Open source | OCR API |
| Price | 2 credits / page | 14 credits / page |
| Per 1,000 pages | $2.00 | $14.00 |
| Typical speed | 1.5 s / page | 3.0 s / page |
| Best for | Chinese and English PDFs | Layout-aware chunks for RAG |
| Inputs | PDF, images | PDF, images, DOCX |
| Languages | Chinese, English | Multilingual |
| Max pages per request | 500 | 1,000 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| Math | ✓ Yes | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | No | ✓ 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 MonkeyOCR if
- Your documents are English or Simplified Chinese PDFs, and pro-3B reports an OmniDocBench overall edit distance of 0.138 (EN) / 0.206 (ZH), the best among the models its authors compared.
- You parse academic papers with formulas and tables, and the output includes LaTeX, Markdown, JSON and block coordinates.
- You have a single consumer GPU: 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.
Choose Chunkr if
- You need a hosted API as well as a self-host option, since Chunkr offers both a cloud service and an AGPL open-source version.
- Your inputs are PDF, PPT, Word and images, and you want semantic chunking for RAG with HTML, Markdown, JSON and bounding box output.
- You want to plug in any OpenAI-compatible LLM endpoint and run via Docker on GPU, CPU-only or Mac ARM.
Key differences
| MonkeyOCR | Chunkr | |
|---|---|---|
| Licence | Code Apache-2.0; v1 weights research and non-commercial only, commercial use needs a written license | AGPL-3.0 or commercial license |
| How you run it | Self-hosted only, via LMDeploy or Docker | Cloud API or self-hosted via Docker (GPU, CPU-only, Mac ARM) |
| Price | Free for research; commercial use needs a separate license from the authors | Hosted credit-based plans (prices not published); open-source version is free but you pay for compute and an LLM endpoint |
| Input formats | PDFs; photographed text and handwriting are not fully supported | PDF, PPT, Word and images; Excel not supported in the open-source version |
| Benchmark | OmniDocBench pro-3B: edit distance 0.138 EN / 0.206 ZH, table TEDS 81.5 / 87.5 (reported by the authors) | Not published |
| Languages | English and Simplified Chinese; Traditional Chinese and multilingual text are not fully supported | Not published |
| Known failure | Pro-3B runs under 1 page/s (about 0.49 pages/s on a 3090); Docker on 20/30/40-series and V100 needs a patched image to avoid an LMDeploy shared-memory error | Open-source version uses community models, less accurate than the cloud's in-house models, and still needs an external LLM endpoint |
Sources
What each one is
MonkeyOCR
Open-source document parser that splits work into structure, recognition and relation steps.
MonkeyOCR detailsChunkr
Document parsing API that segments and chunks pages for RAG. Self-hostable version available.
Chunkr 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": "monkeyocr/monkeyocr","fallbacks": ["chunkr/chunkr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
MonkeyOCR alternativesChunkr alternativesBest OCR models in 2026