Overview

The Flask Live Transcription Starter is an innovative demo application that harnesses the power of Deepgram's API to provide real-time speech-to-text transcription. With a Python Flask backend and a sleek web frontend, this application makes it easy to convert live audio streams into text seamlessly. It's an ideal tool for developers looking to integrate live transcription capabilities into their applications with minimal set-up.

Once set up, users can effortlessly feed audio streams via a WebSocket connection, allowing for immediate transcription results that are dynamically displayed on the client-side. Whether used in production or for testing in development mode, it's a robust solution for anyone interested in leveraging speech recognition technology.

Features

Live Audio Streaming: This application accepts audio stream URLs through a WebSocket connection, facilitating real-time interaction.
Real-time Transcription: It efficiently sends binary audio data to Deepgram's live Speech-to-Text API, providing immediate transcription results.
Flask Backend: The backend is powered by Flask, implementing a straightforward /stt/transcribe endpoint designed in compliance with the STT API contract.
Vite-Powered Frontend: The frontend features a user-friendly interface built with Vite, ensuring a rapid and responsive web experience.
Consistent UI: Utilizing the Deepgram design system, this application offers a consistent aesthetic across various starter applications.
Easy Setup: The quick start guide, including dependency installation and API key configuration, simplifies the setup process for users.
Community Support: Users can engage with the Deepgram community through GitHub Discussions or Discord for assistance, feedback, and contribution opportunities.

Deepgram Starters

Flask Live Transcription

Nextjs Live Transcription