Mathpix Convert vs dots.ocr
Pick Mathpix Convert for math-heavy papers and textbooks that must come out as LaTeX or DOCX, and for handwritten equations, if a hosted API at a per-page price suits you. Pick dots.ocr for multilingual PDFs with mixed tables and text when you can run a 3B model on your own GPU and want MIT-licensed code and no per-page fee. The main split is a paid hosted STEM API versus a free self-hosted layout parser.
Mathpix Convertmathpix/convert | dots.ocrrednote/dots-ocr | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 7 credits / page | 2 credits / page |
| Per 1,000 pages | $7.00 | $2.00 |
| Typical speed | 1.5 s / page | 1.5 s / page |
| Best for | Math and scientific papers | Multilingual document parsing |
| Inputs | PDF, images | PDF, images |
| Languages | 30+ languages | 100+ languages |
| Max pages per request | 500 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | ✓ Yes | ✓ 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 Mathpix Convert if
- You need whole PDFs of math, tables and chemistry exported directly to LaTeX or DOCX.
- You have handwritten equations to convert.
- You want a hosted API with no GPU to run, at $0.005 per PDF page or $0.002 per image.
Choose dots.ocr if
- You want to self-host under MIT-licensed code and pay only for compute.
- You need layout boxes, categories and text from one model, with tables as HTML and formulas as LaTeX, across multiple languages.
- You want published benchmark numbers: 79.1 on olmOCR-bench for the original model and 83.9 for dots.mocr (author-reported).
Key differences
| Mathpix Convert | dots.ocr | |
|---|---|---|
| How you run it | Hosted API | Self-hosted via vLLM or Hugging Face Transformers; a single CUDA GPU with vLLM is the documented path |
| Licence | Proprietary | MIT, with a separate dots.ocr license agreement file in the repository |
| Price | PDF $0.005 per page (0-1M), $0.0035 above 1M; images $0.002 each; one-time $19.99 API key activation fee | Free; you pay for compute |
| Output formats | LaTeX, Markdown, DOCX, HTML, JSON | JSON, Markdown, HTML, LaTeX, bounding boxes, SVG |
| Benchmarks | Not published on general OCR leaderboards | olmOCR-bench 79.1 (dots.ocr) and 83.9 (dots.mocr); OmniDocBench v1.5 text edit distance 0.031 for dots.mocr; all author-reported |
| Known failure | Images with more than 12 text rows may be billed at the PDF page rate | 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 |
| Complex tables and formulas | Built for math, tables and chemistry | Authors say complex tables and formulas remain difficult at this model size |
Sources
- Mathpix billing docs
- Mathpix OCR API
- dots.ocr GitHub README
- dots.ocr blog / limitations
- Docling issue #5146
What each one is
Mathpix Convert
Specialist for math, science papers and handwritten formulas. Outputs LaTeX.
Mathpix Convert 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": "mathpix/convert","fallbacks": ["rednote/dots-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Mathpix Convert alternativesdots.ocr alternativesBest OCR models in 2026