lokutor 1.1.36

Official Python SDK for Lokutor - Real-time AI voice and TTS

Lokutor Python SDK

The official Python SDK for Lokutor - High-performance, real-time AI voice and TTS. Seamlessly transcribe speech, generate intelligent responses, and synthesize natural speech—all through a unified, production-grade interface.

Features

• 🎤 Real-time Voice Input - Automatic microphone recording with low-latency voice activity detection
• 🔊 Natural Speech Output - High-quality text-to-speech synthesis (44.1kHz)
• 🧠 Intelligent Conversations - LLM-powered responses with context awareness
• ⚡ Streaming TTS - Ultra-fast synthesis with raw binary WebSocket streaming
• 🛡️ Production-Ready - Connects directly to Lokutor's secure cloud infrastructure
• 📚 Simple API - Go from code to conversation in 3 lines

Installation

pip install lokutor

Or from source (for development):

git clone https://github.com/lokutor-ai/lokutor-sdk
cd sdk/python
pip install -e .

Requires PyAudio and websocket-client.

Quick Start

Voice Conversation (Start Speaking!)

from lokutor import VoiceAgentClient
from lokutor.config import VoiceStyle, Language

client = VoiceAgentClient(api_key="your-api-key")
client.start_conversation()

Standalone Text-to-Speech

from lokutor import TTSClient, VoiceStyle

client = TTSClient(api_key="your-api-key")
client.synthesize(
    text="Hello world, this is a test of Lokutor streaming TTS.",
    voice=VoiceStyle.F1,
    play=True
)

Core APIs

1. VoiceAgentClient

Full conversational interface that handles VAD, STT, LLM, and TTS orchestration.

client = VoiceAgentClient(
    api_key="your-key",
    prompt="You are a helpful customer service agent",
    voice=VoiceStyle.F2,
    language=Language.ENGLISH
)
client.start_conversation()

Callbacks:

• on_transcription(text): Called when your speech is transcribed.
• on_response(text): Called when the agent's text response is ready.
• on_error(error): Called when a connection or server error occurs.

2. TTSClient

For converting text to high-quality streaming audio.

client = TTSClient(api_key="your-key")
client.synthesize(
    text="Welcome to our platform",
    voice=VoiceStyle.M2,
    play=True,
    block=True
)

Configuration

Voice Styles

Female Voices	Male Voices
VoiceStyle.F1	VoiceStyle.M1
VoiceStyle.F2	VoiceStyle.M2
VoiceStyle.F3	VoiceStyle.M3
VoiceStyle.F4	VoiceStyle.M4
VoiceStyle.F5	VoiceStyle.M5

Supported Languages

• Language.ENGLISH
• Language.SPANISH
• Language.FRENCH
• Language.PORTUGUESE
• Language.KOREAN

Usage Examples

Full Conversation with Callbacks

from lokutor import VoiceAgentClient, VoiceStyle

def handle_user_text(text):
    print(f"User said: {text}")

def handle_agent_text(text):
    print(f"Agent replied: {text}")

client = VoiceAgentClient(
    api_key="demo",
    prompt="You are a friendly pirate assistant",
    on_transcription=handle_user_text,
    on_response=handle_agent_text
)

client.start_conversation()

Scripted TTS with Visemes

from lokutor import TTSClient

def handle_visemes(visemes):
    # Perfect for syncing 3D character lip-sync
    print(f"Syncing lips: {visemes}")

client = TTSClient(api_key="demo")
client.synthesize(
    text="Look at my lips moving!",
    visemes=True,
    on_visemes=handle_visemes
)

Troubleshooting

Connection Issues

The SDK connects to wss://api.lokutor.com. Ensure you have an active internet connection and a valid API key. If you encounter authentication errors, double-check your credentials in the Lokutor Dashboard.

PyAudio Installation

If you encounter issues installing pyaudio, you may need to install portaudio first:

• macOS: brew install portaudio
• Linux: sudo apt-get install python3-pyaudio or sudo apt-get install portaudio19-dev

License

MIT License