Speech-to-Speech Translation API
Overview
Palabra's API enables real-time speech translation over one of two transports:
- WebRTC — best for client applications (browsers, mobile apps). Handled by LiveKit.
- WebSockets — convenient for server-side integrations.
With WebSockets you connect to the Palabra translation server directly, authenticating
with your API Key — a streaming session is created for you automatically.
(Optionally, you can create a session yourself and connect
with its publisher token instead.)
With WebRTC you first create a secure session to get the connection URL and access token.
After connecting, you publish your original audio stream and set your source and target languages.
Palabra then transcribes, translates, and synthesizes your speech into the target languages — all in real time. The translated audio is streamed back to you with minimal latency, ready to play in your application.
How to start
Consider using the Palabra API Python client for WS-based integrations and Palabra API JavaScript client for WebRTC-based integrations.
Read the following articles to start working with the API:
Quick Start Guide
You can also follow the Quick Start Guide, which provides step-by-step instructions and code examples for setting up real-time speech translation.
API Clients
Instead of implementing the low-level API flow yourself, you can use the Palabra API clients: