OCRRouter

Unlimited OCR vs olmOCR

Pick Unlimited OCR to parse long reports and books in one request, with bounding boxes, under an MIT licence. Pick olmOCR for large-scale PDF to Markdown, academic papers with equations, and pages that mix print and handwriting. The main split is page span versus transcription quality: Unlimited OCR reads 40+ pages in one pass, while olmOCR works on a 7B model with a published 82.4 on olmOCR-Bench.

Unlimited OCRbaidu/unlimited-ocrolmOCRallenai/olmocr
TypeOpen sourceOpen source
Price2 credits / page2 credits / page
Per 1,000 pages$2.00$2.00
Typical speed1.2 s / page1.5 s / page
Best forBooks and very long documentsBulk PDF to text for datasets
InputsPDF, imagesPDF, images
Languages30+ languagesMainly English
Max pages per request1,0001,000
Tables✓ Yes✓ Yes
HandwritingNo✓ Yes
Math✓ Yes✓ Yes
Bounding boxes✓ YesNo
ConfidenceNoNo
FormsNoNo
Multilingual✓ YesNo
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 with roughly constant memory, so a long report or book goes in as one request.
  • It returns block types and bounding boxes alongside text, which suits layout-aware extraction.
  • The MIT licence allows commercial use with no copyleft.

Choose olmOCR if

  • It handles equations, tables and handwriting, and strips headers and footers in natural reading order, which fits academic papers and mixed printed and handwritten pages.
  • Self-hosting costs under $200 per million pages per Ai2, and a hosted olmOCR-2-7B is on DeepInfra at $0.09 input / $0.19 output per million tokens.
  • Weights, training data and code are all released, and it scores 82.4 on olmOCR-Bench, above Marker (76.1), MinerU 2.5.4 (75.2) and Mistral OCR (72.0).

Key differences

Unlimited OCRolmOCR
Model size3B total (about 500M active per token, per third-party coverage); Unlimited OCR 2 is 3.34B7B
LicenceMITApache-2.0 (toolkit)
How you run itSelf-host with vLLM, SGLang or transformers; a hosted endpoint is offered in Baidu Cloud's OCR productSelf-host with vLLM on a GPU, or call third-party hosted endpoints
PriceFree self-hosted (you pay for compute); hosted pricing not publishedUnder $200 per million pages self-hosted; DeepInfra $0.09 input / $0.19 output per million tokens
Output formatsMarkdown, plain text, bounding boxesMarkdown, JSON
BenchmarkParseBench mean 46.17 (independent); Unlimited OCR 2 scores 84.6 on MPDocBench and 77.1 on MarathonDocBench (authors)82.4 on olmOCR-Bench v0.4.0 (authors)
Known failureRepetition and looping on some pages, reported hallucinated text, bbox coordinate loops under SGLang on GB10 GPUs, and context capped at 32,768 tokensGets stuck on the last few pages, unreliable image bounding box coordinates, and 38 olmOCR-Bench tests reportedly cannot be passed by a faithful transcription
GPU memoryNVIDIA GPU with CUDA 12.9 or 13.0; version 2 under SGLang runs out of memory on 12 GB GPUsNVIDIA GPU with at least 12 GB VRAM and 30 GB free disk; one report of out of memory on a 12 GB GPU

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

olmOCR

Open-source vision model from Ai2 (often searched as olm ocr) that turns PDFs into clean, reading-order text. Latest release: olmOCR 2.

olmOCR 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": ["allenai/olmocr"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

Unlimited OCR alternativesolmOCR alternativesBest OCR models in 2026