Use the vision capabilities of GPT-4o

Have a look at the documentation for more information on setting up the app.

This demo application uses the latest GPT-4o to analyze images and generate output based on your prompt.
Images are uploaded to Cloudflare R2 storage, a unique URL is generated and is then passed to GPT-4o.
Response is stored in the generations table in Supabase and linked to the user for easy access.
The main frontend logic is found in the app/(apps)/vision folder. You'll find the prompt, configuration file and JSON schema there.
The main configuration file can be found in app/(apps)/vision/toolConfig.ts file.
The API endpoint and logic can be found in app/api/(apps)/vision/route.ts.
Try it out and generate an image!

Your Vision Generations

	Subtitle	Date	Model
No results found.

2026 — built with 🤍 by Kazim