Published Python package

Live Transcriber

Bring live speech into a Python application.

Live Transcriber streams microphone audio to Deepgram and passes final transcript chunks to your code. Use a callback to update an interface or process the text, and optionally save a transcript file.

GitHub · PyPI

What I built

A Python interface around live microphone transcription, with configurable Deepgram recognition options.
Synchronous and asynchronous transcript callbacks, plus optional UTF-8 file output.
File-output pause/resume and graceful shutdown. Pausing file output leaves callbacks running.

A small integration

from livetranscriber import LiveTranscriber

transcriber = LiveTranscriber(callback=print)
transcriber.run()

Requires Python 3.11, microphone access and a Deepgram API key. The source README covers installation and configuration.

Explore the behavior

The walkthrough illustrates callbacks and file output with sample text; it does not start a live transcription session.

Open the walkthrough →