OCRRouter

Nougat vs Tesseract.js

Pick Nougat for English scientific papers where you need LaTeX math and Markdown, and the use is non-commercial. Pick Tesseract.js for in-browser OCR of clean printed text in many languages, or when images must stay on the user's device. The main split is document type and licence: Nougat's weights are CC-BY-NC, while Tesseract.js is Apache-2.0.

Nougatmeta/nougatTesseract.jstesseract/tesseract-js
TypeOpen sourceOpen source
Price2 credits / page1 credit / page
Per 1,000 pages$2.00$1.00
Typical speed3.0 s / page2.0 s / page
Best forarXiv-style papersOCR in the browser
InputsPDF, imagesimages
LanguagesMainly English100+ languages
Max pages per request200500
Tables✓ YesNo
HandwritingNoNo
Math✓ YesNo
Bounding boxesNo✓ Yes
ConfidenceNo✓ Yes
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 Nougat if

  • Your input is English arXiv-style papers, since it was trained on arXiv and PubMed Central documents.
  • You need equations and tables as LaTeX inside Mathpix-compatible Markdown.
  • Your use is non-commercial and you can run a Python package on your own compute.

Choose Tesseract.js if

  • You need OCR to run fully client side in the browser, so images never leave the user's device.
  • You need more than 100 languages, including non-Latin scripts that Nougat does not handle.
  • You need a permissive Apache-2.0 licence for a commercial product, installed through npm, ESM or a CDN script tag.

Key differences

NougatTesseract.js
LicenceCode MIT; weights CC-BY-NC, so no commercial useApache-2.0
How you run itSelf-hosted Python package with a CLI and API; small and base checkpointsJavaScript and WebAssembly library, in the browser or Node.js
PriceFree for non-commercial use; you pay for computeFree; runs on the user's device or your server
Output formatsMarkdown and LaTeXPlain text, hOCR and bounding boxes
LanguagesWorks best on English; README says Chinese, Russian and Japanese do not workMore than 100 languages via Tesseract traineddata files
Input and best documentsScientific PDF pages; math-heavy papers. Not for forms, invoices or business documentsImages such as clean screenshots and scans. Does not accept PDF files; the README points to Scribe.js
Known failureRepetition loops on 1.5% of test pages in the paper, more on out-of-domain documents; looping pages can be dropped as MISSING_PAGELower accuracy than the native Tesseract CLI on the same image; fails on screen resolution text unless images are upscaled first

Sources

What each one is

Nougat

Meta's model that converts academic PDFs to Markdown with LaTeX math.

Nougat details

Tesseract.js

Tesseract compiled to JavaScript and WebAssembly, so OCR runs in the browser or in Node.js.

Tesseract.js 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 10 files at onceNo file?

Use one, fall back to the other

{"model": "meta/nougat","fallbacks": ["tesseract/tesseract-js"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

Nougat alternativesTesseract.js alternativesBest OCR models in 2026