OCRRouter

Nougat vs Phi-4-multimodal

Pick Nougat only for non-commercial work on English arXiv-style papers where you want LaTeX math and tables in Markdown. Pick Phi-4-multimodal for commercial English document Q&A on modest GPUs, because its MIT licence allows commercial use and Nougat's CC-BY-NC weights do not.

Nougatmeta/nougatPhi-4-multimodalmicrosoft/phi-4-multimodal
TypeOpen sourceVision model
Price2 credits / page2 credits / page
Per 1,000 pages$2.00$2.00
Typical speed3.0 s / page2.5 s / page
Best forarXiv-style papersSmall model for OCR and questions
InputsPDF, imagesimages
LanguagesMainly EnglishEnglish for vision
Max pages per request200500
Tables✓ Yes✓ Yes
HandwritingNoNo
Math✓ Yes✓ Yes
Bounding boxesNoNo
ConfidenceNoNo
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 Nougat if

  • Your input is English scientific papers, since it was trained on arXiv and PubMed Central documents.
  • You need LaTeX for equations and tables, written as Mathpix-compatible .mmd Markdown.
  • Your use is non-commercial and you want a small or base checkpoint with a simple CLI and API.

Choose Phi-4-multimodal if

  • You are building a commercial product, because the weights are MIT licensed.
  • You want to ask questions about English documents, where Microsoft reports DocVQA 93.2 and OCRBench 84.4 at 5.6B parameters.
  • You can self-host or use Azure AI Foundry, and you want a 128K token context and audio input in the same model.

Key differences

NougatPhi-4-multimodal
LicenceCode MIT; weights CC-BY-NC, so no commercial useMIT, permissive for commercial use
How you run itSelf-hosted Python package onlySelf-host or use Azure AI Foundry
Output formatsMarkdown and LaTeX (Mathpix-style .mmd)Plain text, Markdown and JSON
BenchmarksNot publishedDocVQA 93.2, OCRBench 84.4, ChartQA 81.4, InfoVQA 72.7 (reported by the authors)
LanguagesWorks best on English; Chinese, Russian and Japanese do not workVision supported for English only; performance drops in other languages
Known failureRepetition loops on 1.5% of test pages in the paper, more on out-of-domain documents; pages can be dropped as MISSING_PAGERepetitive or inconsistent output in long chats; needs eager attention on V100 and older GPUs
Input typesScientific PDF pagesText, image and audio

Sources

What each one is

Nougat

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

Nougat details

Phi-4-multimodal

Microsoft's 5.6B text, vision and audio model (February 2025). Scores 84.4 on OCRBench per its model card.

Phi-4-multimodal 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": "meta/nougat","fallbacks": ["microsoft/phi-4-multimodal"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

Nougat alternativesPhi-4-multimodal alternativesBest OCR models in 2026