From the model card
Fields below are copied from the tags and counters on the HuggingFace repository dots-studio/dots.ocr at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.
- Publisher (HF namespace)
- dots-studio
- Pipeline tag
- image-text-to-text
- Library
- Transformers
- Weight formats
- safetensors
- License tag
mit— read the license file in the repo before relying on it- Language tags
- multilingual; English (en), Chinese (zh)
- Downloads (HF counter at last fetch)
- 358,798
- Likes (HF counter at last fetch)
- 1,321
- Model card
- https://huggingface.co/dots-studio/dots.ocr
Use cases
- Digitizing scanned documents with table and formula preservation
- Extracting structured data from PDF pages or screenshots
- Building document parsing pipelines that handle mixed text and figures
- OCR on historical documents requiring layout-aware transcription
Pros
- Handles tables, formulas, and layout — not just plain text extraction
- 1,318 likes confirms strong community validation of output quality
- Purpose-built for document parsing rather than general VL tasks
- image-to-text and document-parse tags signal task-specific optimization
Cons
- Specialized architecture (dots_ocr) requires custom inference code
- Accuracy on degraded or handwritten input is not benchmarked publicly
- Formula parsing quality depends on notation and typesetting conventions
- Custom model format may not integrate cleanly with generic VL serving stacks
Tags
dots_ocrsafetensorstext-generationimage-to-textocrdocument-parselayouttableformulatransformerscustom_codeimage-text-to-textconversationalenzhmultilinguallicense:miteval-resultsregion:us