Lokutor Python SDK

Lokutor Hero

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

Metadata

Release files for lokutor 1.3.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for lokutor 1.3.0
File	Size	Uploaded
lokutor-1.3.0.tar.gz	18.2 kB	Sep 18, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for lokutor 1.3.0
File	Interpreter	ABI	Platform	Reset
lokutor-1.3.0-py3-none-any.whl 17.1 kB Sep 18, 2026	Python 3	none	any	Details

Total release size: 35.3 kB

Release files / lokutor-1.3.0.tar.gz

Download URL	lokutor-1.3.0.tar.gz
Size	18.2 kB
Tags	Source
SHA-256 checksum How to use checksums	`3a3b85b0cf10043815c63b4e0988a0f365e92f040cb350f2d6b35550f6f933bd`
BLAKE2b-256 checksum How to use checksums	`0dbf5aa88511431dd8895015f92e204f9feb3d62376c264e4f98a051a2ef775a`
Upload date	Sep 18, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.3`

Release files / lokutor-1.3.0-py3-none-any.whl

Download URL	lokutor-1.3.0-py3-none-any.whl
Size	17.1 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`e32d56cbcb27e6f82415d36295abbbe159106d04fbd2c147b27bc384fc320ab0`
BLAKE2b-256 checksum How to use checksums	`4043e0602461a74bfe79a95e346fc41e9efc248b884fb3715454cba2f552d2b9`
Upload date	Sep 18, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`twine/6.2.0 CPython/3.12.3`