OCRRouter

Mathpix Convert vs Qianfan OCR

Pick Mathpix Convert for math-heavy papers and textbooks, handwritten equations, and whole-PDF export to LaTeX or DOCX. Pick Qianfan OCR for self-hosted multilingual document parsing, key information extraction, and tables and formulas under an Apache-2.0 licence. The main split is a hosted, proprietary, per-page API built for STEM against an open 4B model you run yourself.

Mathpix Convertmathpix/convertQianfan OCRbaidu/qianfan-ocr
TypeOCR APIOCR API
Price7 credits / page2 credits / page
Per 1,000 pages$7.00$2.00
Typical speed1.5 s / page2.0 s / page
Best forMath and scientific papersChinese documents
InputsPDF, imagesPDF, images
Languages30+ languagesChinese, English and more
Max pages per request500200
Tables✓ Yes✓ Yes
Handwriting✓ Yes✓ Yes
Math✓ Yes✓ Yes
Bounding boxes✓ YesNo
Confidence✓ YesNo
FormsNoNo
MultilingualNo✓ Yes
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 and turns math, tables and chemistry into LaTeX and Markdown.
  • It exports whole PDFs directly to DOCX and LaTeX, which Qianfan OCR does not list as a DOCX output.
  • It is a hosted API with published per-page prices, so you run no GPU: PDFs cost $0.005 per page, or $0.0035 above 1M pages.

Choose Qianfan OCR if

  • The weights are Apache-2.0, so you can self-host and use it commercially, and you pay only for compute.
  • It covers 192 languages, handwriting and scene text, and scores 93.12 on OmniDocBench v1.5 (reported by the authors).
  • It returns bounding boxes and JSON as well as Markdown, HTML and LaTeX, which suits key information extraction.

Key differences

Mathpix ConvertQianfan OCR
LicenceProprietaryApache-2.0, open weights on Hugging Face
How you run itHosted API onlySelf-host the weights or use the Baidu Qianfan platform
PricePDF $0.005 per page (0-1M), $0.0035 above 1M; images $0.002 each; one-time $19.99 API key feeFree to self-host, you pay for compute; hosted access via Baidu Qianfan platform
Output formatsLaTeX, Markdown, DOCX, HTML, JSONMarkdown, JSON, HTML, LaTeX, bounding boxes
BenchmarksNo published accuracy benchmark on general OCR leaderboardsOmniDocBench v1.5 93.12, olmOCR-Bench 79.8, OCRBench 880 (author-reported)
Best-fit documentsMath-heavy papers and textbooks, handwritten equationsMultilingual documents, key information extraction, tables and formulas
Known drawbacksImages with more than 12 text rows may be billed at the PDF page rate; billed monthly in arrears1.024 pages per second on an A100 even with W8A8; thinking mode can emit up to 16,384 tokens; not for CPU-only deployment

Sources

What each one is

Mathpix Convert

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

Mathpix Convert details

Qianfan OCR

Baidu's OCR-tuned vision model. Strong on Chinese documents and tables.

Qianfan OCR 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": ["baidu/qianfan-ocr"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

Mathpix Convert alternativesQianfan OCR alternativesBest OCR models in 2026