Document OCR Toolkit
Extracts text, tables, and structured data from any document or scanned image.
by zai-org · ★ 476 on GitHub
What it does
You point Claude at an image, PDF, or scanned document. Claude uses the installed GLM-OCR tool to parse it and returns the content as structured data and readable text. It works for invoices, whiteboard photos, forms, and other everyday documents.
Who it’s for
A good fit for anyone working with Word, PDF, Excel or PowerPoint files and people who work with spreadsheets and numbers.
Do you need to code?
Some. To get the most from this one you may need to be comfortable with code, the command line or technical setup.
How to install
Add it to Claude as a skill. Download the skill folder from its source page, zip it, and upload it in Claude’s settings under Skills. See the step-by-step guide.
Safety note
Requires installing a Python package and setting an API key in a terminal; needs a ZhiPu GLM API key.
As with any add-on, only install skills from sources you trust, and read what it does first.
License
Apache-2.0