Skip to main content
The Lokutor SDK handles WebSocket connections, audio I/O, resampling, and conversation state. Available for JavaScript and Python.

​
Installation

Works in browsers (Vite, React, vanilla) and Node.js (16+). Browser apps use the Web Audio API.

​
Get an API key

Sign up at app.lokutor.com and create an API key.

​
Two APIs

Text-to-Speech

Stream synthesized speech from text.

Voice Agent

Full-duplex conversations with microphone and speaker.

Tool Calling

Give your agent tools to interact with external systems.

Visemes

Lip-sync animation with real-time mouth shape data.

​
Platform notes

  • Browser: AudioContext and microphone require a user gesture (button click). See Browser Setup.
  • Node.js: Requires speaker and node-record-lpcm16 for audio I/O. See Node.js Setup.
  • Python: Audio I/O uses pyaudio under the hood (installed automatically as a dependency).