Getting started with LEXO

LEXO (Local EXtraction and OCR) is a desktop app and command-line tool that turns PDFs and images into editable text. It is built for Burmese (Myanmar script) first: Burmese is the default OCR language, text is normalized to Unicode NFC, and a Myanmar font ships with the app so proofreading works on any machine.

Two ways to get text out

LEXO picks the right tool for each page:

• Text layer. Digital PDFs, the kind exported from Word or InDesign, already contain text. LEXO copies it directly. This is instant, lossless, and needs no network or account.
• Google Docs OCR. Scanned PDFs and images have no text, only pictures of text. LEXO renders each page to an image and sends it to Google Docs OCR through your own Google account, then brings the text back.

You can mix both in one document. LEXO routes page by page.

What you need

	Text layer and PDF tools	Google Docs OCR
Python 3.11 or newer	Yes	Yes
Internet connection	No	Yes
Google account	No	Yes
One-time Google Cloud setup	No	Yes

The Google Cloud setup takes six steps and is covered in Google Cloud OAuth setup.

The shortest path

1. Install LEXO:

   uv tool install lexo
   

2. Try it on a digital PDF. No account needed:

   lexo extract report.pdf -o report.txt
   

3. For scans, finish the Google Cloud setup, then sign in once:

   lexo login
   

4. Open the desktop app and work through a document:

   lexo gui
   

   Open a file, tune the pages, pick Text layer or Google Docs OCR, run, proofread, and export. The desktop app guide walks through each panel.

Where to go next

• OCR workflows explains when to use each mode and what to do with legacy-font PDFs.
• Batch OCR covers folders of PDFs.
• Privacy and authentication lists exactly what leaves your computer.