Skip to content

Command Palette

Search for a command to run...

✨ AI Tools

AI that runs in your browser

Every AI model here runs directly in your browser — using ONNX Runtime via the FileCraft Engine. Your documents never get sent to any server for AI inference.

ONNX RuntimeTransformers.js0 bytes to cloud AI

How browser-native AI works

1

Model downloads once

The ONNX model is downloaded directly from Hugging Face CDN and cached in your browser's IndexedDB. Subsequent uses are instant.

2

Runs in a Web Worker

The model runs in a background thread via the FileCraft Engine, so your browser stays responsive while inference happens.

3

Result returns to you

The model output (summary, extracted data) is returned to the page. Nothing is sent anywhere — it's all local.