This open-source website provides a simple interface for modifying images using text-based instructions. You can upload an image, provide a text prompt describing how to change that image, and generate new images based on the prompt.

The model is hosted on Replicate, which exposes a cloud API for running predictions. This website is built with Next.js and hosted on Vercel, and uses Replicate's API to run the Kontext Pro model. The source code is publicly available on GitHub. Pull requests welcome!