docTR vs dots.ocr
Pick docTR if you need word-level text with coordinates from invoices or receipts and want to train your own models in Python. Pick dots.ocr if you need whole documents converted to Markdown with layout, tables and formulas in one pass. The main split is output: docTR returns words and lines only, while dots.ocr returns structured documents but needs a CUDA GPU.
docTRmindee/doctr | dots.ocrrednote/dots-ocr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 1 credit / page | 2 credits / page |
| Per 1,000 pages | $1.00 | $2.00 |
| Typical speed | 0.9 s / page | 1.5 s / page |
| Best for | Word-level OCR with positions | Multilingual document parsing |
| Inputs | PDF, images | PDF, images |
| Languages | Mainly Latin scripts | 100+ languages |
| Max pages per request | 500 | 500 |
| Tables | No | ✓ Yes |
| Handwriting | No | No |
| Math | No | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | ✓ Yes | 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 docTR if
- You want nested page, block, line and word JSON with coordinates, and you will build the extraction logic yourself.
- You want a choice of detection and recognition architectures, with straight or rotated boxes, and a key information extraction predictor.
- You need to run on CPU or NVIDIA GPU with ready Docker images, under Apache-2.0.
Choose dots.ocr if
- You need multilingual PDFs converted to Markdown, with tables as HTML and formulas as LaTeX.
- You work with academic or business documents that mix tables and text, and want layout boxes and categories from a single model.
- You want chart-to-SVG conversion through the dots.mocr-svg variant, under MIT-licensed code.
Key differences
| docTR | dots.ocr | |
|---|---|---|
| What it is | Two-stage text detection and recognition library from Mindee, on PyTorch | 3B parameter document parsing VLM (1.7B LLM backbone) from rednote hi lab |
| Licence | Apache-2.0 | MIT, plus a separate dots.ocr license agreement file in the repository |
| How you run it | Local Python library, with Docker images | Self-hosted via vLLM or Hugging Face Transformers |
| Hardware | CPU or NVIDIA GPU; GPU Docker images need CUDA 12.2 or newer on the host | Single CUDA GPU with vLLM is the documented path; no VRAM figure stated; listed as not suited to CPU-only production |
| Output formats | JSON, plain text, bounding boxes | JSON, Markdown, HTML, LaTeX, bounding boxes, SVG |
| Tables and formulas | No table, formula or Markdown output | Tables as HTML and formulas as LaTeX, but complex tables and formulas remain difficult |
| Benchmarks | Not published | olmOCR-bench 79.1 for dots.ocr and 83.9 for dots.mocr (1.5), as reported by the authors |
| Known failure | Cropping images leads to poorer results (issue #2007); split-crop merging misjoins repeated characters (issue #2229) | Runs of ellipses or underscores can make output repeat endlessly; a table above 1000 grid cells raises an error and drops the page in Docling (issue #5146) |
Sources
- docTR GitHub README
- Issue #2007
- Issue #2229
- dots.ocr GitHub README
- dots.ocr blog / limitations
- Docling issue #5146
What each one is
docTR
Open-source OCR library from Mindee with separate text detection and recognition models.
docTR detailsdots.ocr
Single vision model from rednote that does layout detection and content recognition together.
dots.ocr 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": "mindee/doctr","fallbacks": ["rednote/dots-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
docTR alternativesdots.ocr alternativesBest OCR models in 2026