I am a heavy user of Google Assistant. Multiple Google speakers (Nest Audio) are synchronized throughout my home, playing music and executing my voice commands daily. Numerous electronic devices in my house are configured to be controlled via Google.

Then came the release of ChatGPT. Its natural language processing capabilities made the so-called ‘AI speakers’ I had been using seem rather primitive. This piqued my curiosity about using such advanced natural language processing for a voice assistant.

The GPT Voice Assistant I created leverages OpenAI’s ChatGPT as an AI voice assistant. It allows you to converse with GPT in various languages (Korean, English, etc.) through your PC’s microphone. It is written in Python, with speech recognition (STT) and speech synthesis (TTS) handled by Google’s API.

The project is managed on Github, and detailed usage instructions can be found on the WIKI.

Currently, ChatGPT itself offers voice conversation features. Therefore, there is no longer a need to use this project.