Docling vs Chunkr
Pick Docling if you ingest born-digital PDFs and Office files into a RAG pipeline and want MIT-licensed code that runs locally with lossless JSON. Pick Chunkr if you want layout parsing with OCR bounding boxes and semantic chunking, either as a hosted API or self-hosted behind an OpenAI-compatible LLM endpoint. The main split is format coverage and licence: Docling reads many more formats under MIT, while Chunkr is AGPL-3.0 and needs a commercial licence for closed use.
Doclingibm/docling | Chunkrchunkr/chunkr | |
|---|---|---|
| Type | Open source | OCR API |
| Price | 2 credits / page | 14 credits / page |
| Per 1,000 pages | $2.00 | $14.00 |
| Typical speed | 2.2 s / page | 3.0 s / page |
| Best for | Structured export for RAG | Layout-aware chunks for RAG |
| Inputs | PDF, images, DOCX | PDF, images, DOCX |
| Languages | Latin scripts, CJK | 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 Docling if
- You need one API for PDF, DOCX, PPTX, XLSX, HTML, EPUB, images, emails and XBRL.
- You want MIT-licensed code with no commercial licence to negotiate for closed-source products.
- You want lossless JSON that keeps layout, reading order and table structure, and a choice of OCR engines (EasyOCR, Tesseract, RapidOCR) or the 258M GraniteDocling VLM.
Choose Chunkr if
- You want OCR output with bounding boxes and semantic chunking of PDFs and slides.
- You want the option of a hosted cloud API, or a Docker self-host with GPU, CPU-only and Mac ARM options.
- You already run an OpenAI-compatible LLM endpoint and want layout parsing that works with it.
Key differences
| Docling | Chunkr | |
|---|---|---|
| Licence | MIT (individual models have their own licences) | AGPL-3.0 or commercial licence; closed use needs the commercial one |
| How you run it | Python library that runs locally on macOS, Linux and Windows, CPU or GPU | Cloud API, or self-host via Docker with GPU, CPU-only and Mac ARM options |
| Price | Free, self-hosted; you pay for compute | Hosted credit-based plans, prices not published; open-source version is free but you pay for compute and an LLM endpoint |
| Input formats | PDF, DOCX, PPTX, XLSX, HTML, EPUB, images, emails, XBRL | PDF, PPT, Word and images; Excel is not supported in the open-source version |
| Output formats | Markdown, HTML, JSON, DocTags, plain text | HTML, Markdown, JSON, bounding boxes |
| Scanned-page accuracy | Depends on the OCR engine you plug in, not Docling itself | Open-source version uses community models, less accurate than the cloud's in-house models |
| Known failure | Version 2.143.1 drops words and numbers with full-page Tesseract OCR (issue #5254); reading order can merge neighbouring columns (issue #5132) | Self-host still needs an external LLM endpoint |
Sources
- Docling GitHub README
- Supported formats
- Issue #5254 OCR regression
- Issue #5132 reading order
- Chunkr GitHub README
- Chunkr site
What each one is
Chunkr
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": "ibm/docling","fallbacks": ["chunkr/chunkr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Docling alternativesChunkr alternativesBest OCR models in 2026