Image to prompt for ChatGPT
ChatGPT creates images from plain language. Upload a picture and get a clear description of what is in it and how it looks, ready to paste as your request.
How your ChatGPT prompt is written
Natural language
The universal prompt is written in full sentences, which is how ChatGPT image generation reads best.
Ask for the format
Add the aspect ratio in words, like square or wide, at the end of the prompt. ChatGPT picks it up from there.
Iterate in the chat
Use Remix to change one attribute, then paste the new version as a follow up message in the same conversation.
Example ChatGPT prompt
A cozy reading nook, described the way ChatGPT reads it.
A cozy reading nook by a large rainy window, a deep green velvet armchair with a knitted blanket, a stack of old books and a steaming mug on a small wooden table. Warm lamp light mixes with cool gray daylight. Photographed at eye level with a 35mm lens, soft and calm mood, wide format.
How to get a ChatGPT prompt from an image
- 01
Upload your image
JPEG, PNG or WebP. It is resized in your browser first.
- 02
We read the image
Subject, style, lighting, lens, palette and visible text are pulled out.
- 03
Copy the ChatGPT prompt
The ChatGPT tab opens first. Copy it and paste it into ChatGPT.
ChatGPT prompt questions
Can ChatGPT turn an image into a prompt by itself?
It can describe an image, but PromptXray is built for this job: it names the lens, light setup and palette, and gives you versions for Midjourney, Flux and Nano Banana from the same upload.
Does it work with other models too?
Yes. The universal prompt works with most modern image models, including Ideogram, Leonardo and Firefly.
Does ChatGPT see my original image?
No. Your image is sent once to the vision model that writes the prompt and is not saved on our servers. ChatGPT only gets the text you paste.