DIY AI APPS · OFFLINE TRANSCRIBER
Your words.
On your device.
Turn English recordings into editable text and subtitles. Prepare your browser once, then transcribe without an internet connection.
Start transcribing Open a workspace to bookmark- 01
Prepare once
Download and verify the speech model while online.
- 02
Add your audio
Choose a recording or record a voice note locally.
- 03
Review and keep
Correct your transcript, then save it to your device.
Make your next recording useful.
No audio is uploaded. Microphone recording and local recovery drafts are optional.
What works offline?
After downloading the model, open the bookmarked workspace, choose or record English audio, edit transcripts and export files without a connection. The full website needs an internet connection; bookmark the separate workspace for offline use.
Will this work on my device?
Start with an up-to-date desktop Chrome or Edge browser over HTTPS. Allow at least 250 MB of free storage plus memory for your recording, and test a short sample before relying on it. Mobile performance has not been verified. This version uses English Whisper Tiny; it does not identify speakers or translate audio.
Where is my work saved?
Audio stays in memory and is cleared when you close the workspace. You can explicitly save text and timestamps to this browser, enable a recovery draft, or download an editable project. Browser data can be cleared, so download copies you want to keep. During initial setup, Hugging Face and jsDelivr receive ordinary connection information, never your recording.