Voice based interface to ChatGPT
The project, built in node, aims to be a platform agnostic application to interact with ChatGPT using your voice, like a natural conversation.
The end goal is to provide an easy to install solution for a RaspberryPI based assistant.
This package aims to minimize its reliance on external packages as much as possible.
At this stage this project can only run on MacOS, more adapters will come soon. The project requires node 18+
Create a .env
file on the root of the project with the contents in .env.example
Then run yarn install
and yarn start
to launch.
More options can be found in src/constants.js
The speech-to-text supports the following engines:
The functionality requires the following packages to be installed:
The text-to-speech supports the following engines:
Requires an API key and a voice ID set in the .env file.
Text-to-speech uses the native say cli interface
- ✅ MVP
- ✅ OSX adapters
- 🟨 Eleven labs adapters
- 🟨 pico voice adapters
- 🟨 Move constants to env file
- 🟨 Expose web UI
- 🟨 Install instructions for Raspberry PI