Chrome AI is a demo project that runs large language models locally within the browser. It showcases the capabilities of modern web technologies to leverage advanced AI functionalities directly in the browser environment without needing server-side processing.
- Local Execution: Runs large language models directly in the browser.
- Real-time Processing: Provides instant responses and interactions.
- Privacy-focused: Processes data locally, ensuring user privacy and data security.
- Latest version of Google Chrome (127+) or any compatible Chromium-based browser.
- Install Chrome Canary: Ensure you have version 127. Download Chrome Canary.
- Enable Prompt API: Open
chrome://flags/#prompt-api-for-gemini-nano
, set it to "Enabled". - Enable Optimization Guide: Open
chrome://flags/#optimization-guide-on-device-model
, set it to "Enabled BypassPerfRequirement". Restart the browser. - Download Model: Go to
chrome://components/
, find "Optimization Guide On Device Model", ensure itβs fully downloaded. If the version is "0.0.0.0", click "Check for update". - Troubleshoot: If the "Optimization Guide On Device Model" is not displayed, disable the settings in steps 2 and 3, restart your browser and re-enable it.
- Verify Setup: Open a webpage, press F12, and check
window.ai
in the console.
Test Code:
const model = await window.ai.createTextSession();
await model.prompt("Who are you?");
Clone the repository:
git clone https://github.com/yourusername/chromeai.git
cd chromeai
pnpm i
pnpm dev
Open localhost:3000
in your browser to start using the AI.
We welcome contributions! Please fork the repository and submit pull requests. For major changes, please open an issue first to discuss what you would like to change.
This project is licensed under the MIT License - see the LICENSE file for details.
For any questions or feedback, please contact Lightning Joyce on Twitter.