OCRRouter

Jina OCR vs Unlimited OCR

Pick Unlimited OCR for commercial products and for long reports or books parsed in one request. It is MIT licensed and reads 40+ pages in one pass, and it also returns bounding boxes. Pick Jina OCR for non-commercial or research PDF to Markdown, where it has the stronger published scores, but commercial use needs a license from Jina/Elastic.

Jina OCRjina/jina-ocrUnlimited OCRbaidu/unlimited-ocr
TypeOpen sourceOpen source
Price1 credit / page2 credits / page
Per 1,000 pages$1.00$2.00
Typical speed1.5 s / page1.2 s / page
Best forClean Markdown at low costBooks and very long documents
InputsPDF, imagesPDF, images
LanguagesAbout 100 languages30+ languages
Max pages per request2001,000
Tables✓ Yes✓ Yes
HandwritingNoNo
Math✓ Yes✓ Yes
Bounding boxesNo✓ Yes
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 Jina OCR if

  • Its authors report 91.14 on OmniDocBench v1.6 and 83.4 on olmOCR-Bench, 7.4 points above DeepSeek-OCR.
  • It has about 570M active parameters and runs at 2.57 pages per second on an A100.
  • It outputs HTML and LaTeX as well as Markdown, and it falls back to CPU in Transformers if you have no GPU.

Choose Unlimited OCR if

  • The MIT license allows commercial use with no copyleft, while Jina OCR is CC-BY-NC-4.0.
  • It parses 40+ pages in one forward pass with roughly constant memory, and Baidu reports 77.1 on MarathonDocBench versus 33.7 for the next model.
  • It returns block types and bounding boxes with the text, which suits layout-aware extraction and bulk PDF to text for RAG.

Key differences

Jina OCRUnlimited OCR
LicenceCC-BY-NC-4.0; commercial use needs a license from Jina/ElasticMIT; commercial use allowed
How you run itSelf-host, or use Jina's hosted APISelf-host with vLLM, SGLang or transformers, or use the hosted endpoint in Baidu Cloud's OCR product
Output formatsMarkdown, HTML, LaTeXMarkdown, plain text, bounding boxes
HardwareGPU recommended (A100, L4 benchmarked); CPU fallback in TransformersNVIDIA GPU with CUDA 12.9 or 13.0; Unlimited-OCR-2 under SGLang runs out of memory on 12 GB GPUs; not for CPU-only machines
Multi-page inputPage-to-Markdown model; no multi-page limit is40+ pages in one pass, but context is capped at 32,768 tokens and multi-page input only works in base mode
Published benchmarksOmniDocBench v1.6 91.14 and olmOCR-Bench 83.4 (authors)ParseBench mean 46.17 (independent); MPDocBench 84.6 and MarathonDocBench 77.1 for Unlimited OCR 2 (authors)
Known failureHosted API returns HTTP 503 on cold start; retry after 30 to 60 seconds. Reported scores use a stricter prompt than the default.Text repetition and looping (issue #55), reported hallucinated text (issue #58), and little bold or italic formatting kept (ParseBench text formatting 0.97)

Sources

What each one is

Jina OCR

Compact page-to-Markdown model from Jina AI, built on the DeepSeek-OCR encoder.

Jina OCR details

Unlimited OCR

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

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

Questions

More comparisons

Jina OCR alternativesUnlimited OCR alternativesBest OCR models in 2026