AI client for Chrome‑integrated Gemini Nano

A minimal client that runs prompts against Gemini Nano via Chrome's Prompt API.

Gemini Nano setup (most important)

• Supported platforms: Windows 10/11, macOS 13+, or Linux (desktop Chrome only; not Android/iOS/ChromeOS).
• Storage: ≈ 22 GB free on the Chrome profile volume for the model.
• GPU: > 4 GB VRAM recommended.
• Network: Unmetered connection for first download.

1. In Chrome, open chrome://flags and enable:

   • #prompt-api-for-gemini-nano
   • #optimization-guide-on-device-model ("Enabled" or "Enabled BypassPerfRequirement") Then relaunch Chrome.

2. Open chrome://components, find "Optimization Guide On Device Model", click "Check for update" and wait for the model to download.

3. Verify model status in chrome://on-device-internals (Model status + File path).

See the official Prompt API docs: AI on Chrome — Prompt API.

Develop

• Install: npm i
• Run: npm run dev

Troubleshooting

• Ensure desktop Chrome is up to date and meets the hardware/storage requirements above.
• Leave Chrome running during first download; large initial download may take time.
• Check chrome://on-device-internals for model status and file location.