This Streamlit app allows you to test the effectiveness of your prompts using OpenAI's text models: gpt-4, gpt-3.5-turbo, and text-davinci-003. Different models may behave slightly differently to the same prompt, and this tool allows you to compare how the models would behave in response to your prompt.
To get started, enter an initial prompt and, optionally, a follow-up message. You can also adjust the model parameters to fine-tune your tests. You can keep adding additional follow-up messages to have test conversations with the models.
You can access the app on the Streamlit Cloud community at gpt-compare.streamlit.app.
To use the app, you will need an OpenAI API key. Don't have one yet? Create one on the OpenAI webiste. Once you have your API key, enter it into the app when prompted.
If you have any feedback or questions about this app, please reach out to me on Twitter at @dclin.
Thank you for checking out the tool!