TSClient - Talkscriber Python Client

A comprehensive Python client library for Talkscriber's Live Transcription and Text-to-Speech services. This package provides easy-to-use APIs for real-time speech-to-text transcription and ultra-low latency text-to-speech conversion.

Features

Live Transcription

• Real-time speech-to-text transcription via WebSocket
• Support for 50+ languages
• Smart turn detection using ML models
• Translation capabilities
• File and microphone input support

Text-to-Speech

• Ultra-low latency streaming (speech starts in <0.1 seconds)
• Real-time audio playback
• Multiple voice options
• Audio file saving
• Configurable buffering for optimal performance

Installation

Quick Install

pip install tsclient

System Dependencies

The package requires some system-level dependencies for audio processing:

macOS:

# Install PortAudio (required for PyAudio)
brew install portaudio

# Then install the package
pip install tsclient

Ubuntu/Debian:

# Install system dependencies
sudo apt-get update
sudo apt-get install libasound2-dev portaudio19-dev python3-dev

# Then install the package
pip install tsclient

Windows:

# No additional setup required - just install the package
pip install tsclient

CentOS/RHEL/Fedora:

# Install system dependencies
sudo yum install portaudio-devel python3-devel
# or for newer versions:
sudo dnf install portaudio-devel python3-devel

# Then install the package
pip install tsclient

Development Install

For development or to run examples:

git clone https://github.com/Talkscriber/ts-client.git
cd ts-client
pip install -e .

Quick Start

Get Your API Key

1. Visit Talkscriber Dashboard
2. Sign up or log in
3. Generate an API key
4. Set it as an environment variable:

export TALKSCRIBER_API_KEY="your_api_key_here"

Live Transcription

from talkscriber.stt import TranscriptionClient

# Initialize client
client = TranscriptionClient(
    host="wss://api.talkscriber.com",
    port=9090,
    api_key="your_api_key",
    language="en"
)

# Start live transcription from microphone
client()

# Or transcribe from file
client("audio_file.wav")

Text-to-Speech

from talkscriber.tts import TalkScriberTTSClient

# Initialize TTS client
tts_client = TalkScriberTTSClient(
    api_key="your_api_key",
    text="Hello, world!",
    speaker_name="tara"
)

# Generate and play speech
tts_client.run_simple_test()

Command Line Interface

The package includes convenient CLI tools:

Live Transcription CLI

# Transcribe from microphone
talkscriber-stt --api-key YOUR_KEY --language en

# Transcribe from file
talkscriber-stt --api-key YOUR_KEY --file audio.wav

# Enable multilingual detection
talkscriber-stt --api-key YOUR_KEY --multilingual

Text-to-Speech CLI

# Basic TTS
talkscriber-tts --api-key YOUR_KEY --text "Hello, world!"

# Save to file
talkscriber-tts --api-key YOUR_KEY --text "Hello" --save output.wav

# Use specific voice
talkscriber-tts --api-key YOUR_KEY --text "Hello" --speaker tara

API Reference

Live Transcription

TranscriptionClient

TranscriptionClient(
    host: str,
    port: int,
    api_key: str,
    multilingual: bool = False,
    language: str = None,
    translate: bool = False,
    enable_turn_detection: bool = False,
    turn_detection_timeout: float = 0.6
)

Parameters:

• host: WebSocket host URL
• port: WebSocket port number
• api_key: Your Talkscriber API key
• multilingual: Enable auto language detection
• language: Language code (e.g., "en", "es", "fr")
• translate: Enable translation
• enable_turn_detection: Enable ML-based turn detection
• turn_detection_timeout: Timeout for turn detection

Text-to-Speech

TalkScriberTTSClient

TalkScriberTTSClient(
    host: str = "api.talkscriber.com",
    port: int = 9099,
    text: str,
    speaker_name: str = "tara",
    api_key: str,
    enable_playback: bool = True,
    save_audio_path: str = None
)

Parameters:

• host: TTS server hostname
• port: TTS server port
• text: Text to convert to speech
• speaker_name: Voice to use
• api_key: Your Talkscriber API key
• enable_playback: Enable real-time audio playback
• save_audio_path: Optional path to save audio file

Supported Languages

The client supports 50+ languages including English, Spanish, French, German, Chinese, Japanese, and many more. See the full language list in our documentation.

Examples

Check out the examples/ directory for comprehensive examples organized by category:

Quick Start

# First, install the package
pip install tsclient

# Then run examples
# Interactive demo (recommended for beginners)
python examples/integration/comprehensive_demo.py

# Basic STT (microphone)
python examples/stt/basic.py

# Basic TTS (text-to-speech)
python examples/tts/basic.py

Example Categories

Speech-to-Text (STT):

• stt/basic.py - Real-time microphone transcription
• stt/file.py - Transcribe audio files
• stt/multilingual.py - Automatic language detection

Text-to-Speech (TTS):

• tts/basic.py - Basic TTS with real-time playback
• tts/save_file.py - Generate and save audio files
• tts/silent.py - Silent TTS generation
• tts/different_voices.py - Multiple voice comparison
• tts/batch.py - Batch processing from text files

Integration:

• integration/stt_to_tts.py - Complete voice processing pipeline
• integration/comprehensive_demo.py - Interactive demo

Configuration:

• configuration/config_patterns.py - Configuration patterns and best practices

See the examples README for detailed usage instructions and the complete directory structure.

Documentation

• Full Documentation
• API Reference
• Language Support

Links

• Talkscriber Dashboard - Get your API key
• Official Website
• GitHub Repository

License

This project is licensed under the MIT License - see the LICENSE.md file for details.