Fridge-photo based recipe generator by using Google AI Studio

This post is my submission for DEV Education Track: Build Apps with Google AI Studio.

What I Built

I built a simple app that gives recipe ideas using a photo of your fridge. The user uploads a fridge photo, and the app finds the ingredients and shows easy recipes.
I used Google AI Studio with the Vision model. My prompt tells Gemini to detect ingredients and suggest simple recipes from what it sees.

This is the prompt that I used:

You are a helpful cooking assistant.

The user will upload a photo of the inside of their fridge. First, >analyze the image and identify only the visible food ingredients (e.g. >eggs, milk, vegetables, sauces, fruits, leftovers, etc.). Ignore non->food items like containers, packaging, bottles, or background objects.

Then, based on the identified ingredients, suggest 1 or 2 simple >recipes that can be made using only those ingredients.

Each recipe should include:

• Recipe name
• List of required ingredients (from the image)
• Easy step-by-step instructions (maximum 4 steps)

If you cannot find enough ingredients, reply politely:

"I couldn't find enough ingredients to suggest a recipe. Please try a clearer photo."

Keep your response short, friendly, and beginner-friendly.

Demo

Link to my app (Google AI Studio App)
Live Demo

My Experience

I really enjoyed building this app. It was easy to use Google AI Studio and create something useful with just a prompt.

I learned how to write a good prompt for image input and how AI can turn images into smart responses. I plan to add more features to this app later!