How to Build a Python Voice Assistant with Google GenAI & SpeechRecognition

Build Your Own Voice-Controlled AI Assistant with Python, Gemini, and Speech Tools

Building a voice assistant used to require complex audio processing pipelines, machine learning model hosting, and tedious configuration files. Today, combining modern generative AI with lightweight audio packages makes it possible to build a fully functional, talking, listening AI assistant in under 50 lines of Python code.

The command pip install google-genai speechrecognition pyttsx3 pyaudio unlocks the ultimate toolkit for building local voice-activated AI agents.

What Each Package Does

Before diving into code, let’s break down what this powerhouse quartet brings to your project:

• google-genai: The official, modern Google GenAI SDK that connects your application to Google’s state-of-the-art Gemini models (like gemini-2.5-flash) for intelligent text generation and reasoning.
• speechrecognition: A wrapper library that simplifies speech-to-text conversion, hooking seamlessly into web APIs (like Google Web Speech API) to translate microphone audio into strings.
• pyaudio: The underlying engine that binds Python to your operating system’s audio input/output, allowing speechrecognition to capture real-time sound streams from your microphone.
• pyttsx3: A cross-platform text-to-speech library that works completely offline, giving your Python scripts a natural voice to speak answers back to you.

Step 1: Installation

Open your terminal or command prompt and run the installation command:

pip install google-genai speechrecognition pyttsx3 pyaudio

Platform Note for PyAudio: If you encounter errors during pyaudio installation on Linux or macOS, you may need to install system audio drivers first (