VoiceChat-py

A real-time voice chat application that uses Groq's API for speech-to-text, language model processing, and text-to-speech capabilities. Supports both Groq and SmallestAI for TTS.

Overview

VoiceChat-py is a Python-based CLI tool that enables seamless voice conversations with AI. It allows you to:

1. Record audio by holding a key (default: spacebar)
2. Transcribe the audio using Groq's Whisper API
3. Process the transcription with a language model (LLM)
4. Convert the LLM's response back to speech (Groq or SmallestAI)
5. Play the audio response (with cancel option)

Features

• Real-time audio recording: Record audio while holding down a key (default: spacebar)
• Audio level visualization: See live audio input levels in your terminal
• Speech-to-text: Transcribe audio using Groq's Whisper API
• Language model processing: Process transcriptions with Groq's LLM API
• Text-to-speech: Convert LLM responses to speech using Groq's TTS API or SmallestAI
• Configurable: Customize models, providers, and parameters via environment variables
• Flexible TTS provider: Choose between Groq and SmallestAI for text-to-speech
• Audio playback with cancel: Cancel playback at any time with a key (default: Esc)
• Clear error handling: Helpful messages if dependencies or API keys are missing

Requirements

• Python 3.7+
• Groq API key (required for all modes)
• (Optional) SmallestAI API key (for SmallestAI TTS)
• Audio input/output devices (microphone and speakers)
• The following Python packages:
  • numpy
  • pyaudio
  • keyboard
  • groq
  • python-dotenv
  • pydub
  • playsound3

Install all dependencies with:

pip install -r requirements.txt

Installation

1. Clone the repository:

   git clone https://github.com/vegeta03/voicechat-py.git
   cd voicechat-py

2. Install required packages:

   pip install -r requirements.txt

3. Copy the sample environment file and edit it:

   cp .env.sample .env

   Then edit .env to add your actual API keys and configuration.

Usage

Run the main script:

python main.py

• To record audio: Hold down the spacebar (or configured key)
• To stop recording: Release the key
• To cancel recording: Press Esc (or configured cancel key)
• The application will transcribe your audio, send it to the LLM, convert the response to speech, and play back the response.
• To cancel playback: Press Esc during playback

Switching TTS Provider

By default, Groq is used for text-to-speech. To use SmallestAI, set TTS_PROVIDER=smallestai in your .env and provide your SmallestAI API key.

Configuration

All configuration is via environment variables (set in .env).

Variable	Description	Default
GROQ_API_KEY	Your Groq API key (required)
GROQ_STT	Speech-to-text model name	distil-whisper-large-v3-en
GROQ_LLM	LLM model name	llama-3.1-8b-instant
GROQ_TTS	TTS model name (Groq only)	playai-tts
MAX_CHARS_PER_CHUNK	Max chars per TTS chunk	800
TTS_PROVIDER	TTS provider: 'groq' or 'smallestai'	groq
SMALLESTAI_API_KEY	Your SmallestAI API key (if using SmallestAI TTS)
SMALLESTAI_MODEL	SmallestAI TTS model name	lightning
VOICE_ID	Voice ID (for SmallestAI, e.g., emily)	emily
RECORD_KEY	Key to hold for recording audio	space
CANCEL_KEY	Key to cancel recording/playback	esc

You can add or update these in your .env file.

Project Structure

• main.py — Main script with all functionality
• requirements.txt — Python dependencies
• .env.sample — Sample environment configuration

Troubleshooting & FAQ

Q: I get an ImportError for a missing package.

• A: Install all dependencies with pip install -r requirements.txt.

Q: The script says my API key is missing.

• A: Make sure you have set your GROQ_API_KEY (and SMALLESTAI_API_KEY if needed) in your .env file.

Q: Audio recording/playback doesn't work.

• A: Check your microphone and speaker devices. Make sure they are enabled and accessible.

Q: How do I switch to SmallestAI for TTS?

• A: Set TTS_PROVIDER=smallestai and provide your SMALLESTAI_API_KEY in .env.

Q: Can I use a different LLM or TTS model?

• A: Yes, set GROQ_LLM, GROQ_TTS, or SMALLESTAI_MODEL as desired.

Acknowledgements

• This project uses the Groq API for AI capabilities
• Built with PyAudio for audio recording
• SmallestAI TTS support via SmallestAI (if enabled)