I have a model deployed on Space with Gradio. This makes the model extremely accessible with the exposed inference API.
However, is it recommended if I use the API for in production environment? What are the pros/cons compared to hosting the model on AWS, Azure or GCP? Any thoughts?