Unlimited-OCR: Read text from page images and whole PDFs in one pass
Added · Updated
Made by Baidu
View profile on GitHub (opens in a new tab)Turn photos of documents, multi-page scans and PDFs into text you can copy and search.
For developers who process long documents with their own GPU.
What it does
Unlimited-OCR is a vision-language model from Baidu, meaning it looks at an image and writes text about it. OCR (optical character recognition) means reading the text in a picture. The creators say it aims to push DeepSeek-OCR one step further, with a focus on what they call one-shot long-horizon parsing: reading many pages in a single run instead of one page at a time.
How to use it
Input → action → output: load the model in Python and give it a single page image with a document-parsing prompt, or a list of page images for multi-page parsing. For a PDF, the source's example first converts each page to an image with PyMuPDF. The model writes the parsed results to an output folder you choose.
Common use cases
- Convert a long scanned PDF into page images, run multi-page parsing, and get the document's text saved to a folder.
- Upload a photo of a printed page to the hosted demo and get its text back without installing anything.
- Run it behind an SGLang server and send page images to its OpenAI-compatible API from your own app to get text back.
Can I use this?
Requirements Install the Python packages listed on the model card, load it with trust_remote_code (it runs custom code from the repository), and move it to a CUDA GPU. Docker images exist for vLLM serving. A hosted version is also offered on Baidu Cloud; running the open weights yourself is separate. MIT licensed, an open source license.
Caveat The source mostly covers setup; it gives few details on languages, layouts or accuracy, so test it on your own documents.
Worth knowing The model card is short: supported languages beyond a multilingual tag, accuracy and output format are not documented there. PDFs must first be turned into page images.
Why it's here
Not trending this week
Last seen on the GitHub trending list on 30 October 2026. It stays on BuildTube so you can still find it.
- 30,886 people have starred it on GitHub.
Numbers checked on 8 November 2026; not refreshed since.
Behind it
View model pageBuildTube has not run or verified this project. Everything above is written from what the creator published.
Made this?
Quick questions
- What is Unlimited-OCR?
- Unlimited-OCR is a vision-language model from Baidu, meaning it looks at an image and writes text about it.
- What can I use Unlimited-OCR for?
- Convert a long scanned PDF into page images, run multi-page parsing, and get the document's text saved to a folder; Upload a photo of a printed page to the hosted demo and get its text back without installing anything; Run it behind an SGLang server and send page images to its OpenAI-compatible API from your own app to get text back
- Can I run Unlimited-OCR on my computer?
- It runs from a terminal on your own computer. Install the Python packages listed on the model card, load it with trust_remote_code (it runs custom code from the repository), and move it to a CUDA GPU.
- What do I need to use Unlimited-OCR?
- An NVIDIA GPU with CUDA and Python with Hugging Face Transformers, or a vLLM or SGLang server. Exact GPU memory needs are not documented.
- What license does Unlimited-OCR use?
- MIT licensed, an open source license.
More from this week
- Idol Turn one picture of an object into a 3D Gaussian splat
Idol turns one photo, render or generated image of an object into a 3D Gaussian splat file.
- P-H3-Image-Comfy Make and edit 8-megapixel images in ComfyUI
Turn short text requests into 8-megapixel images and edit pictures on your own graphics card.
- agent-desktop Let AI agents read and click desktop apps
For developers building computer-use agents: agent-desktop lets an AI agent read and operate desktop apps on Mac, Windows or Linux from a command line.
Explore by topic
Post this
Download share imageFor makers
Made this? Show it’s on BuildTube — add the badge to your README or embed the card on your site.
README badge (Markdown)
[](https://buildtube.app/p/gh-baidu-unlimited-ocr/)Embed card (HTML)
<iframe src="https://buildtube.app/embed/gh-baidu-unlimited-ocr/" width="560" height="240" style="border:0;max-width:100%" loading="lazy" title="Unlimited-OCR on BuildTube"></iframe>