Overview
XTTSService provides multilingual voice synthesis through a locally hosted Coqui XTTS-v2 streaming server, speaking with one of the server’s studio speakers.
XTTS API Reference
Pipecat’s API methods for XTTS integration
Example Implementation
Complete example with a studio speaker
XTTS Repository
Official XTTS streaming server repository
Installation
XTTS requires a running streaming server. Start the server using Docker:Prerequisites
XTTS Server Setup
Before usingXTTSService, you need:
- Docker Environment: Set up Docker with GPU support for optimal performance
- XTTS Server: Run the XTTS streaming server container
Required Configuration
- Server URL: Configure the XTTS server endpoint (default:
http://localhost:8000) - Voice Selection: Choose one of the server’s studio speakers
GPU acceleration is recommended for optimal performance. The server requires
CUDA support for best results.
Configuration
XTTSService
Deprecated since 0.0.105, removed in 2.0.0. Use
settings=XTTSService.Settings(voice=...) instead.Deprecated since 0.0.106, removed in 2.0.0. Use
settings=XTTSService.Settings(language=...) instead.Runtime-configurable settings. See Settings below.
Output audio sample rate in Hz. When
None, uses the pipeline’s configured
sample rate. Audio is automatically resampled from XTTS’s native 24kHz output.Settings
Runtime-configurable settings passed via thesettings constructor argument using XTTSService.Settings(...). These can be updated mid-conversation with TTSUpdateSettingsFrame. See Service Settings for details.
Usage
Basic Setup
With Language Configuration
Notes
- Local server required: XTTS requires a locally running streaming server (via Docker). The service connects to this server over HTTP.
- Studio speakers: On startup, the service fetches available “studio speakers” from the server’s
/studio_speakersendpoint. Thevoicesetting must match one of these speakers; otherwise each request pushes anErrorFrame. - Audio resampling: XTTS natively outputs audio at 24kHz. The service automatically resamples to match the pipeline’s configured sample rate.
- GPU recommended: The XTTS server performs best with CUDA-enabled GPU acceleration. CPU inference is significantly slower.
- No API key required: XTTS runs locally, so no external API credentials are needed.