TIL: Handling WebSocket Streams With the websockets Library in Python

Python’s websockets library opens a persistent two way connection. It is the cleanest way to stream audio data to a speech-to-text API.

import asyncio
import websockets

async def listen():
    async with websockets.connect('wss://api.deepgram.com/v1/listen') as ws:
        await ws.send('{type:KeepAlive}')
        msg = await ws.recv()
        print(msg)

asyncio.run(listen())

The async with block handles connect and disconnect. send and recv are coroutines that look like regular function calls.

async def stream_audio(ws, audio_path):
    with open(audio_path, 'rb') as f:
        while chunk := f.read(4096):
            await ws.send(chunk)
    await ws.send('{type:CloseStream}')

Streaming audio works by sending binary chunks through the same WebSocket connection. The API responds with partial transcripts as it processes each chunk.

Does websockets support SSL?

Yes. Use wss:// URLs. The library handles TLS automatically.

What Python version is required?

Python 3.6 or later. The library uses async/await syntax which was stabilised in 3.6.