This is a streamlit app that generates appropriate captions based on the user's image and selected options. You can generate professional-looking or rather catchy captions using this tool. It can handle most of the use cases like selfies, sceneries, group images, brand marketing images, etc. It also has an option to provide additional details.
The app uses Google's Gemini-Pro-Vision multimodal API. Careful prompt engineering has been done to ensure output in a particular format and structure.