Real-time Speech-to-Text in LiveKit with AssemblyAI

This repository contains a minimal example of how to use AssemblyAI's real-time transcription API with LiveKit. See the companion blog How to add real-time Speech-to-Text to your LiveKit application for a full walkthrough.

See the companion blog for more detailed instructions on how to set up and run this example, but here is the general overview:

1. Clone this repository and cd into it
2. Rename .env.example to .env
3. Go to livekit.io and sign up
4. In your LiveKit dashboard, go to Settings > Keys and click on your API key
5. Copy the URL, API key, and secret values into the .env file
6. Go to assemblyai.com and sign up, and then set up billing to get access to Streaming Speech-to-Text
7. Copy your AssemblyAI API key into the .env file
8. pip install -r requirements.txt to install requirements
9. python stt_agent.py to run the Speech-to-Text agent
10. Go the the LiveKit Agents Playground and sign in with your LiveKit account. Begin speaking and you will see your speech transcribed in real-time both in the Playground and in the agent's terminal.