About Gemini OCR
A free, independent OCR app that runs in your browser on Google's Gemini models.
Gemini OCR turns images, PDFs and web pages into text, Markdown and structured data. It is a static web app: there is no Gemini OCR server, no account and no upload to us. Your browser talks to Google's Gemini API directly, using your own API key.
Who makes it
Gemini OCR is built and maintained by Mansour Damanpak (GitHub, LinkedIn). Questions, bugs and ideas: mansoor.damanpak@gmail.com.
The tools
- Simple OCR: one image or PDF to Markdown.
- Templates: invoices, receipts, resumes and business cards to JSON, CSV or Excel.
- Link to Text: up to 20 web pages or online PDFs, fetched live.
- Bulk OCR: up to 50 files in one run.
- Agent OCR: field extraction that re-checks low-confidence fields.
- Markdown Converter: documents to Markdown entirely on your device, no API key.
How it is built
React and TypeScript, hosted on Vercel as static files. Model calls use the official @google/genai SDK; see the Gemini OCR API guide for the exact requests. What happens to your data is described on the privacy page.
Independence
Gemini OCR is an independent project. It is not made, endorsed or supported by Google. Gemini is a trademark of Google LLC.