From the model card
Fields below are copied from the tags and counters on the HuggingFace repository baidu/Unlimited-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)
- baidu
- 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
- Papers cited
- arXiv:2606.23050
- Downloads (HF counter at last fetch)
- 3,008,635
- Likes (HF counter at last fetch)
- 4,182
- Model card
- https://huggingface.co/baidu/Unlimited-OCR
Use cases
- Extracting text from scanned PDFs and mixed-layout documents
- Multilingual OCR across East Asian, Latin, and Arabic scripts
- Receipt and invoice field parsing
- Document digitization pipelines
- Scene text recognition in photos
Pros
- MIT license permits commercial use
- Multilingual coverage across diverse scripts
- Published eval results enable objective comparison
- 1.6M downloads reflects broad community validation
Cons
- custom_code dependency — not plug-and-play with standard VLM pipelines
- Feature-extraction pipeline tag understates complexity; output decoding needs verification
- Inference resource requirements not documented in model card
- No fine-tuning instructions for domain adaptation
Tags
transformerssafetensorsunlimited-ocrfeature-extractionbaiduvision-languageocrcustom_codeimage-text-to-textmultilingualarxiv:2606.23050license:miteval-resultsregion:usdeploy:sagemaker