This project is a Text or Voice Input AI Assistant that uses the OpenAI API to generate responses and the ElevenLabs Text-to-Speech API to convert the responses into audio quickly.
- Real-time speech-to-speech conversation
- Customizable voice selection
- Easy-to-use command-line interface
- Powered by OpenAI and ElevenLabs
- Paid ElevenLabs subscription required
-
Ensure you have Miniconda installed on your system.
-
Get your OpenAI API Key
-
Get your ElevenLabs API Key
-
To run
main_ws.pyyou will need a Azure OpenAI Deployment and Speech Service Resource
- Fork and clone the repository:
git clone https://github.com/<username>/OAI-QuickChat.git
cd OAI-QuickChat- Create a conda environment:
conda create -n qchat python=3.12 -y- Install the required packages:
pip install -r requirements.txt- Create a
.envfile in the project root directory using the provided.env.templatefile. Fill in the required API keys and voice ID:
cp .env.template .env- Edit the
.envfile and add your OpenAI API key, ElevenLabs API key, and desired ElevenLabs voice ID.
- Run the AI Assistant:
python main.py to type your requests in the terminal.
python main.py --voice to talk to the assistant.
- Use the following if you have access to AzureOpenAI and Elevenlabs websockets version of the assistant:
python main_ws.py to type your requests in the terminal.
python main_ws.py --voice to talk to the assistant.
-
Enter your question or command when prompted.
-
The AI Assistant will generate a response and play it as audio.
-
To exit the AI Assistant, press
CTRL+C.
You can customize the voice used by the AI Assistant by changing the ELEVENLABS_VOICE_ID in the .env file.
A link to the pre-made voices is in the .env.template file.
This project is licensed under the MIT License. See the LICENSE file for details.
- Fork the repository.
- Create a new branch for your feature or bugfix.
- Commit your changes and push to your fork.
- Create a pull request.
If you encounter any issues or have questions, please open an issue on GitHub.