OCRRouter

Mathpix Convert vs Granite Docling

Pick Mathpix Convert if you need math-heavy papers or textbooks turned into LaTeX or DOCX, or handwritten equations read, and you accept a paid hosted API. Pick Granite Docling if you want a free, Apache-2.0, self-hosted model for English PDFs with tables, code and equations inside a Docling pipeline. The main split is a proprietary per-page API against a 258M model you run yourself.

Mathpix Convertmathpix/convertGranite Doclingibm/granite-docling
TypeOCR APIOpen source
Price7 credits / page1 credit / page
Per 1,000 pages$7.00$1.00
Typical speed1.5 s / page1.5 s / page
Best forMath and scientific papersSmall-model document conversion
InputsPDF, imagesPDF, images
Languages30+ languagesEnglish, experimental CJK and Arabic
Max pages per request500500
Tables✓ Yes✓ Yes
Handwriting✓ YesNo
Math✓ Yes✓ Yes
Bounding boxes✓ Yes✓ Yes
Confidence✓ YesNo
FormsNoNo
MultilingualNoNo
On OCRRouterAvailableAvailable

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

  • It is built for STEM content, so math, tables and chemistry come out as LaTeX or Mathpix Markdown.
  • It exports whole PDFs straight to DOCX or LaTeX, and it lists handwritten equations as a good fit.
  • It is a hosted API, so you run no GPU and manage no model weights.

Choose Granite Docling if

  • It is Apache-2.0 and free to run, so you pay only for compute, with no per-page fee or activation charge.
  • At 258M parameters it suits CPU or small GPU conversion of English PDFs with tables, code and equations.
  • It outputs DocTags plus Markdown, HTML, LaTeX and bounding boxes, and it fits directly into Docling pipelines.

Key differences

Mathpix ConvertGranite Docling
LicenceProprietaryApache-2.0
How you run itHosted APISelf-hosted, typically inside Docling (258M parameters)
PricePDF $0.005 per page up to 1M pages, then $0.0035; images $0.002 each; one-time $19.99 API key feeFree; you pay for compute
Output formatsLaTeX, Markdown, DOCX, HTML, JSONDocTags, Markdown, HTML, LaTeX, bounding boxes
Published benchmarksNot published on general OCR leaderboardsAuthors report table TEDS 0.97 on FinTabNet, equation F1 0.968, full-page OCR F1 0.84, OCRBench 500
LanguagesNot publishedEnglish primary; Japanese, Arabic and Chinese are experimental
Known failure modesImages with more than 12 text rows may be billed at the PDF page rate; billed monthly in arrearsRepetitive loops on page headings and invoices; exclamation-mark output on GPUs without bfloat16 such as T4 unless run in float32; vLLM wte error needs the untied revision

Sources

What each one is

Mathpix Convert

Specialist for math, science papers and handwritten formulas. Outputs LaTeX.

Mathpix Convert details

Granite Docling

Compact IBM vision model that converts pages to DocTags for the Docling pipeline.

Granite Docling details

Try 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": "mathpix/convert","fallbacks": ["ibm/granite-docling"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

Mathpix Convert alternativesGranite Docling alternativesBest OCR models in 2026