GitHub radar
Pull Text and Tables Out of Any Scanned or Photographed Document
TeleOCR is a compact model that reads documents from photos and scans, even tilted shots, and returns clean text with tables. Set it up once and run any paperwork through it right on your own machine.
It's a document reader: show it a photo or scan of a page and it gives back the text and tables in readable form.
When it helps
Useful when you need to pull text from a photographed document, receipt, or old contract shot at an angle. Not for chatting with an AI agent or writing code.
Pros
- Reads tilted photos of documents without separate page straightening
- About 1GB in size — runs on a regular laptop
- Keeps tables and reading order intact on multi-column pages
- Scores higher than much larger general models on document parsing
Cons
- Doesn't chat or write code — only parses documents
- Needs a laptop with at least 8GB of RAM
How to set it up — step by step
- 1Open your AI agent (Claude Code, Codex) in your project folder.
- 2Send the agent the text from the block below.
- 3Wait while the agent installs the model and sets up a way to run it on your machine.
- 4Give the agent a scan or photo of the document you want to read.
- 5Check that the text and tables in the reply match the original document.
Text for your agent
Copy this and send it to your agent — Claude Code, Codex, any of them:
Set up the document-parsing model TeleOCR from this page: https://huggingface.co/XingChen-AGI/TeleOCR — figure out how to run it on my machine and show me how to feed it a scanned document or photo and get back text and tables.
A regular laptop with 8 GB of RAM or more — the model itself takes up about 1 GB.
Open on Hugging Face▌ More finds