Making the web accessible for everyone using Chrome's built-in AIβcompletely offline and private.
An AI-powered Chrome extension that brings comprehensive web accessibility features to your browser: summarize articles, translate languages, simplify complex text, and proofread grammar. All powered by Chrome's built-in AI APIs with Gemini Nano, running entirely on-device for complete privacy.
Extract key points from long articles in seconds using the Summarizer API
- One-click summarization of entire webpages
- Bullet-point format for quick comprehension
- Perfect for research, news, and time-saving
Translate text into 20+ languages with the Translator API
- Spanish, French, German, Chinese, Japanese, Arabic, and more
- Select any text and get instant translations
- Works completely offline after initial setup
- No data sent to external servers
Transform complex content into easy-to-understand language using the Rewriter API
- Converts technical jargon to simple vocabulary
- Maintains original meaning while improving readability
- Ideal for students, non-native speakers, and accessibility
Correct spelling, grammar, and punctuation with the Proofreader API
- Professional-grade corrections
- Instant feedback for writing improvement
- Perfect for drafting responses and comments
AI-generated descriptions of images using the Prompt API
- Right-click any image for natural language descriptions
- Accessibility compliance with alt-text generation
- Multimodal understanding of visual content
- π 100% Private: All AI processing happens on your deviceβno data ever leaves your computer
- β‘ Works Offline: Once AI models download, everything works without internet
- π Accessible: Helps millions including students, researchers, non-native speakers, and people with disabilities
- π Fast: Near-instant results after initial model download
- π Open Source: MIT licensed, fully auditable codebase
- Chrome Canary (version 128+) or Chrome Dev Channel
- Chrome Built-in AI Early Preview Program enrollment
- Experimental flags enabled
-
Download Chrome Canary
https://www.google.com/chrome/canary/ -
Join Chrome Built-in AI Early Preview Program
https://developer.chrome.com/docs/ai/join-epp -
Enable Required Flags
- Navigate to:
chrome://flags/#enable-experimental-web-platform-features - Set to: Enabled
- Navigate to:
chrome://flags/#optimization-guide-on-device-model - Set to: Enabled BypassPerfRequirement
- Restart Chrome completely
- Navigate to:
-
Install Extension
- Go to
chrome://extensions/ - Enable Developer mode (toggle in top right)
- Click "Load unpacked"
- Select this extension folder
- Extension icon should appear in toolbar
- Go to
-
First Use
- On first feature use, Gemini Nano model will download (~10-60 seconds)
- Progress displayed in UI
- Subsequent uses are instant
- Navigate to any article or webpage
- Click the extension icon in toolbar
- Click "Summarize Page" button
- Key points appear in floating overlay (bottom-right)
- Select any text on webpage
- Click extension icon
- Choose source and target languages
- Click "Translate" button
- Translation appears in overlay
- Select complex text
- Click extension icon
- Click "Simplify Text" button
- Simplified version appears in overlay
- Select text to check
- Click extension icon
- Click "Proofread Text" button
- Corrected version appears in overlay
- Right-click on any image
- Select "Analyze Image with AI"
- Description appears in notification
This extension showcases 5 different Chrome Built-in AI APIs:
| API | Purpose | Implementation |
|---|---|---|
| Summarizer | Extract key points | type: 'key-points' |
| Translator | Cross-language translation | 20+ languages supported |
| Rewriter | Text simplification | tone: 'more-casual', length: 'shorter' |
| Proofreader | Grammar correction | outputLanguage: 'en' |
| Prompt | Multimodal image analysis | Planned feature |
- Manifest V3 for modern Chrome extensions
- Content Scripts for page interaction
- Page MAIN World execution for AI API access
- Chrome Message Passing for component communication
- On-device Processing with Gemini Nano
- β All AI runs locally on your device
- β No data transmitted to external servers
- β No tracking or analytics
- β Works offline after model download
- β Open source code (MIT License)
- Summarization: Visit any news article β Click "Summarize Page"
- Translation: Select Spanish text β Translate to English
- Simplification: Select technical paragraph β Click "Simplify Text"
- Proofreading: Select text with errors β Click "Proofread Text"
- Offline Test: Disconnect internet β Repeat above (should still work)
- News: CNN.com, BBC.com
- Foreign Language: ElPais.com (Spanish)
- Technical: Wikipedia articles
- Any webpage with text content
- Ensure you've enabled both required flags (see Installation step 3)
- Restart Chrome completely (quit and reopen)
- Verify Chrome Canary version is 128+
- Check enrollment in Early Preview Program
- Normal! Gemini Nano model downloads on first use (10-60 seconds)
- Download happens once per API
- Watch for progress indicator in UI
- Subsequent uses are near-instant
- Check browser console for specific errors (
F12β Console tab) - Verify extension is enabled in
chrome://extensions/ - Reload extension after enabling flags
- Ensure you've clicked buttons (user activation required)
See detailed troubleshooting in the repository issues section.
- Lines of Code: ~400 JavaScript
- APIs Used: 5 Chrome Built-in AI APIs
- AI Model: Gemini Nano (on-device)
- Privacy: 100% local processing
- Offline: β Full functionality
- License: MIT (Open Source)
Contributions welcome! This is an open-source project built for the Google Chrome Built-in AI Challenge 2025.
- Fork this repository
- Create feature branch (
git checkout -b feature/AmazingFeature) - Commit changes (
git commit -m 'Add AmazingFeature') - Push to branch (
git push origin feature/AmazingFeature) - Open Pull Request
We actively encourage feedback! For the Most Valuable Feedback Prize, please test the extension and complete this form:
Chrome AI Feedback Form: https://forms.gle/V3QzcVcNMotTiYdd9
Your feedback helps improve Chrome's AI APIs for the entire developer community.
This project is licensed under the MIT License - see the LICENSE file for details.
Built for the Google Chrome Built-in AI Challenge 2025
Category: Most Helpful - Chrome Extension
- Google Chrome Team for Built-in AI APIs
- Chrome Built-in AI Early Preview Program
- Gemini Nano team for on-device AI capabilities
Made with β€οΈ for web accessibility