English-only · local in browser

Speech to text without sending audio to a server.

Upload an audio file or record a live session. Whisper runs on the client device and keeps transcription inside the browser.

Model

Loading Whisper base.en…

Runtime

Preparing local inference

Upload audio

MP3, WAV, M4A, WebM

Best for prerecorded files. Audio is decoded in the browser and passed to Whisper locally.

Select or drop an audio file No file selected

Microphone

Record locally in the browser. When you stop the session, the captured audio is transcribed on-device.

Microphone idle

Waiting for the Whisper model to finish loading.