Welcome to KugelAudio
KugelAudio documentation: a text-to-speech API for voice agents and real-time apps, with SDKs, WebSocket streaming, voice cloning and framework integrations.
What is KugelAudio?
KugelAudio is a text-to-speech (TTS) platform for real-time applications such as voice agents, interactive applications, and content creation tools.
To make your first request, create an API key in the dashboard, then follow the Quick Start. Every new account starts with free Text to Speech credit, no credit card needed. Rates and plans are on the pricing page.
Integrations
Building a voice agent on a framework? Our official plugins add KugelAudio as its TTS provider.
LiveKit
Official TTS plugin for LiveKit Agents in Python and TypeScript
Pipecat
Official TTS service for Pipecat pipelines in Python
Vapi
Custom TTS provider for Vapi assistants
Cognigy
Speech vendor for Cognigy Voice Gateway
ElevenLabs compatibility
Use any ElevenLabs-compatible SDK or tool
Choose your path
Make your first request
Install an SDK or use cURL and generate audio
Build a voice agent
Stream LLM output to speech, or plug into LiveKit, Pipecat, Vapi, or Cognigy
Use the API without an SDK
REST and WebSocket wire protocol
Run on your infrastructure
On-premise deployment for Enterprise customers
Voices
Browse voices and create custom clones
Transcribe audio
Speech to text in ten languages with Kugel Ear 1
Hold a live conversation
Speech to speech with Kugel Fluid 1 over one WebSocket
Account, limits and help
Credits, top-ups and invoices
How the balance works, invoices, and what to do when a payment fails
Rate limits and tiers
The limits per tier and how to reach the next one
Data handling
What is stored, for how long, and the data processing agreement
Status and support
Check the status page, or write to hello@kugelaudio.com
SDKs
Python
Python SDK quickstart
TypeScript/JavaScript
TypeScript/JavaScript SDK quickstart
Java
Java SDK quickstart
Rust
Rust SDK quickstart
C++
C++ SDK quickstart
Dart/Flutter
Dart and Flutter SDK quickstart
Key features
Two current models
Use kugel-3.5, the default and newest model, or kugel-3. See Models for their capabilities.
Natural pauses with break tags
Insert pauses with <break> tags. See Breaks.
WebSocket streaming
Stream audio chunks as they are generated. Suited to LLM integrations where text arrives token by token. See Streaming.
Voice cloning
Create custom voices from clean reference audio. See Voice cloning for recording requirements.
Multi-language support
One multilingual model for 39 languages, including DE, EN, FR, ES, IT, NL, PT, PL, RU, ZH, JA, KO, and AR. See supported languages for the full list.