OCRmyPDF vs docTR
Pick OCRmyPDF to make scanned PDFs searchable and keep them as PDF/A, for example in an office scanner pipeline or an archive. Pick docTR when you need word-level text with coordinates as JSON, for example to feed invoices or receipts into your own extraction code. Neither returns tables or Markdown.
OCRmyPDFocrmypdf/ocrmypdf | docTRmindee/doctr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 1 credit / page | 1 credit / page |
| Per 1,000 pages | $1.00 | $1.00 |
| Typical speed | 1.0 s / page | 0.9 s / page |
| Best for | Searchable PDFs from scans | Word-level OCR with positions |
| Inputs | PDF, images | PDF, images |
| Languages | 100+ languages (via Tesseract) | Mainly Latin scripts |
| Max pages per request | 1,000 | 500 |
| Tables | No | No |
| Handwriting | No | No |
| Math | No | No |
| Bounding boxes | No | ✓ Yes |
| Confidence | No | ✓ Yes |
| 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 OCRmyPDF if
- You need a validated searchable PDF/A as the output, which it produces by default.
- You want the OCR layer added without re-encoding the rest of the PDF where possible.
- You need 100+ languages through Tesseract language packs and want to use all CPU cores with no GPU.
Choose docTR if
- You need a nested page, block, line and word structure with coordinates as JSON.
- You want to choose detection and recognition architectures, with straight or rotated boxes, or train your own models.
- You want Apache-2.0 licensing, ready Docker images, and a key information extraction predictor.
Key differences
| OCRmyPDF | docTR | |
|---|---|---|
| Licence | MPL-2.0 | Apache-2.0 |
| Output formats | Searchable PDF (PDF/A by default) and plain text | JSON, plain text and bounding boxes |
| Engine and hardware | Tesseract; CPU only, spread across all cores | Two-stage deep learning detection and recognition on PyTorch; CPU or NVIDIA GPU (GPU Docker images need CUDA 12.2 or newer on the host) |
| Languages | 100+ through Tesseract language packs | Not published |
| Layout, tables, Markdown | None; accuracy is capped by Tesseract | None; returns words and lines only |
| Price | Free, self-hosted (you pay for compute) | Free, self-hosted (you pay for compute) |
| Known failure | --skip-text with default PDF/A output removes form fields (issue #1790); PDFs whose forms share one Resources dictionary can hang (issue #1763) | Split-crop merging can misjoin readings across repeated characters (issue #2229); cropping images leads to poorer results (issue #2007) |
Sources
What each one is
OCRmyPDF
Command-line tool that adds a searchable text layer to scanned PDFs, using Tesseract.
OCRmyPDF detailsdocTR
Open-source OCR library from Mindee with separate text detection and recognition models.
docTR detailsTry both on your own file
Run the file once with each model and compare the output.
PDF, PNG, JPG, WebP, HEIC. Up to 20 MB each, 10 files.No file?
Use one, fall back to the other
{"model": "ocrmypdf/ocrmypdf","fallbacks": ["mindee/doctr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
OCRmyPDF vs KrakenOCRmyPDF vs RapidOCRRapidOCR vs docTREasyOCR vs docTRdocTR vs OpenOCRdocTR vs Qwen-VLGLM-OCR vs docTRdocTR vs Phi-4-multimodal
OCRmyPDF alternativesdocTR alternativesBest OCR models in 2026