v2.6.0 Released

Give your AI agents
a voice that feels real.

Production-grade text-to-speech engine for Python. Zero-cost abstractions, compiled control graphs, and swappable backends.

$ pip install voice-soundboard
quickstart.py
from voice_soundboard import VoiceEngine # Start the engine engine = VoiceEngine() # Say something naturally result = engine.speak( "Hello world! This is my AI voice.", emotion="excited" ) print(f"Audio saved to: {result.audio_path}")

Zero-Cost Abstractions

Emotions, styles, and SSML compile into a control graph before runtime. The engine stays fast and unaware of high-level concepts.

Swappable Backends

Switch between Kokoro (GPU optimized), Piper (CPU fast), or your own custom TTS engine without changing application code.

Streaming Ready

Built for real-time applications. Stream audio sentence-by-sentence with sentence boundary awareness.