OCRRouter

olmOCR vs GOT-OCR 2.0

Pick olmOCR when you need to turn large batches of PDFs into Markdown and have a GPU with 12 GB or more. Pick GOT-OCR 2.0 when you need a small 580M model that reads formulas, sheet music or chemical structures, or OCR on a chosen region of a page. The main split is scope: olmOCR is built for whole-page PDF to Markdown, GOT-OCR 2.0 is a general reader for many content types.

olmOCRallenai/olmocrGOT-OCR 2.0stepfun/got-ocr-2
TypeOpen sourceOpen source
Price2 credits / page1 credit / page
Per 1,000 pages$2.00$1.00
Typical speed1.5 s / page1.5 s / page
Best forBulk PDF to text for datasetsFormatted text, formulas and charts
InputsPDF, imagesimages
LanguagesMainly EnglishChinese, English
Max pages per request1,000100
Tables✓ Yes✓ Yes
Handwriting✓ YesNo
Math✓ Yes✓ Yes
Bounding boxesNo✓ Yes
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 olmOCR if

  • It scores 82.4 on olmOCR-Bench, above Marker (76.1), MinerU 2.5.4 (75.2) and Mistral OCR (72.0) in the same table, while GOT-OCR 2.0 lists no benchmark scores.
  • Self-hosting costs under $200 per million pages per Ai2, and hosted olmOCR-2-7B on DeepInfra costs $0.09 input / $0.19 output per million tokens.
  • It handles equations, tables and handwriting, and strips headers and footers in natural reading order, which suits academic papers and mixed printed and handwritten pages.

Choose GOT-OCR 2.0 if

  • At 580M parameters it is much smaller than the 7B olmOCR, and a llama.cpp port exists for CPU users.
  • It outputs LaTeX, TikZ, SMILES and kern as well as plain text and Markdown, so it covers formulas, chemical structures and sheet music.
  • Region OCR guided by box coordinates or color lets you read one part of a page, and it supports multi-page OCR and batched inference in Hugging Face Transformers.

Key differences

olmOCRGOT-OCR 2.0
Size7B parameters580M parameters
How you run itSelf-host with vLLM on a GPU, or use third-party hosted endpointsSelf-host only, via Hugging Face Transformers
HardwareNVIDIA GPU with at least 12 GB VRAM and 30 GB free diskCUDA 11.8 and PyTorch 2.0.1 per README; CPU users pointed to a llama.cpp port
LicenceApache-2.0 (toolkit)Code Apache-2.0; data CC BY-NC 4.0
Output formatsMarkdown, JSONPlain text, Markdown, LaTeX, TikZ, SMILES, kern
Benchmark82.4 on olmOCR-Bench (v0.4.0)Not published
Known failureOut of memory on a 12 GB GPU (issue #424); gets stuck on the last few pages (issue #433); unreliable image bounding box coordinates (issue #437)Output can loop characters midway (issue #226); returns wrong text on images with no text (issue #235); format mode is much worse than plain OCR mode (issue #317)

Sources

What each one is

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

GOT-OCR 2.0

Unified OCR model that reads plain text, tables, formulas, sheet music and charts.

GOT-OCR 2.0 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": "allenai/olmocr","fallbacks": ["stepfun/got-ocr-2"],"file_url": "https://example.com/document.pdf"}

Questions

More comparisons

olmOCR alternativesGOT-OCR 2.0 alternativesBest OCR models in 2026