OCRRouter

Unlimited OCR vs OCRmyPDF

Pick Unlimited OCR when you need Markdown, text or bounding boxes from long documents such as reports and books, and you have an NVIDIA GPU. Pick OCRmyPDF when you need to turn scanned PDFs into searchable PDF/A on CPU-only machines. The main split is output: a parsed document versus the same PDF with a text layer added.

Unlimited OCRbaidu/unlimited-ocrOCRmyPDFocrmypdf/ocrmypdf
TypeOpen sourceOpen source
Price2 credits / page1 credit / page
Per 1,000 pages$2.00$1.00
Typical speed1.2 s / page1.0 s / page
Best forBooks and very long documentsSearchable PDFs from scans
InputsPDF, imagesPDF, images
Languages30+ languages100+ languages (via Tesseract)
Max pages per request1,0001,000
Tables✓ YesNo
HandwritingNoNo
Math✓ YesNo
Bounding boxes✓ YesNo
ConfidenceNoNo
FormsNoNo
Multilingual✓ Yes✓ 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 Unlimited OCR if

  • It parses 40+ pages in a single forward pass and keeps memory roughly constant, so long reports and books go through in one request.
  • It outputs Markdown, plain text and bounding boxes with block types, which suits RAG pipelines and layout-aware extraction.
  • It is MIT licensed, so commercial use has no copyleft, and Baidu Cloud offers a hosted endpoint if you do not want to run GPUs.

Choose OCRmyPDF if

  • It produces validated searchable PDF/A by default, which fits archiving and compliance workflows.
  • It runs on CPU only and spreads work across all cores, so it fits office scanner pipelines with no GPU.
  • It uses Tesseract, so language packs cover 100+ languages, and it inserts the OCR layer losslessly where possible without re-encoding other content.

Key differences

Unlimited OCROCRmyPDF
What it producesMarkdown, plain text and bounding boxesSearchable PDF/A and plain text; no layout, table or Markdown output
HardwareNVIDIA GPU with CUDA 12.9 or 13.0; version 2 under SGLang runs out of memory on 12 GB GPUs; not for CPU-only machinesCPU only, all cores used
LicenceMITMPL-2.0
How you run itSelf-host with vLLM, SGLang or transformers; hosted option in Baidu CloudLocal command-line tool
PriceFree self-hosted, you pay for compute; hosted endpoint offered in Baidu Cloud's OCR productFree self-hosted, you pay for compute
BenchmarksParseBench mean 46.17; Unlimited OCR 2 scores 77.1 on MarathonDocBench versus 33.7 for the next model, per BaiduNot published
Known failureText repetition and looping on some pages, reported hallucinated text, bbox coordinate loops in SGLang on GB10 GPUs, and bold or italic styling mostly lostFile size can double or quadruple with --deskew --clean-final, --skip-text with default PDF/A output removes form fields, and PDFs whose forms share one Resources dictionary can hang

Sources

What each one is

Unlimited OCR

Open-source model from Baidu that reads hundreds of pages in one pass with flat memory use.

Unlimited OCR details

OCRmyPDF

Command-line tool that adds a searchable text layer to scanned PDFs, using Tesseract.

OCRmyPDF 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": "baidu/unlimited-ocr","fallbacks": ["ocrmypdf/ocrmypdf"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

Unlimited OCR alternativesOCRmyPDF alternativesBest OCR models in 2026