Skip to main content
MGToolbox

Speech to Text (In-Browser AI)

AIRuns On Your Device

Transcribe microphone audio or audio files to text in English, Spanish, French, Portuguese, or Japanese — AI that runs in your browser, no uploads.

Speech recognition runs on your device

First use downloads a compact speech model, cached afterwards for offline use. Your audio is never uploaded to any server.

— or —
Runs On Your Device
The AI model downloads once and is cached — your files and voice never leave your browser.
Instant Client-Side Speed
Zero network upload delays
Free & Unlimited
No registration or credit card needed

Free Private Speech-to-Text — AI That Runs on Your Device

Transcribe recordings or live microphone audio with a modern speech-recognition model that runs entirely inside your browser using WebGPU (with a WASM fallback). Unlike online transcription services, your audio is never uploaded anywhere — the model weights are downloaded once, cached locally, and the transcription itself happens on your device.

Five languages are supported: English (Moonshine Base), plus Spanish, French, Portuguese, and Japanese via a compact multilingual model (Whisper Tiny). The first use downloads a ~40–90 MB model — after that, it works offline. A genuine alternative to subscription transcription tools for interviews, voice memos, and dictation, especially for anything you would rather not hand to a cloud service.

Frequently Asked Questions

Everything you need to know about using Speech to Text (In-Browser AI) safely and efficiently.

No. The speech model is downloaded to your browser and runs locally. Your audio is processed on your device and never transmitted. (The one-time model download comes from a CDN — that is code, not your data.)
Recommended Next Steps

Related Online Utilities