AI tool, 10 free a day, 30 signed in
Image to prompt
To turn an image into a prompt, open the image here, keep Detailed prompt or pick another style, and press Write the prompt. Our AI describes the subject, setting, lighting, colors and style in words you can paste into an AI image generator. It can also write keyword tags, alt text or a plain description. The image is sent only when you press the button, and we keep neither it nor the words.
- Prompt, tags, alt text or a description
- Sent only when you press Write the prompt
- No sign-up for 10 a day
Credits pay for Good and Max on the image readers and the AI humanizer. Low stays free every day.
- Starter 50 credits No ads for 30 days $4.99
- PlusRecommended 120 credits+20% bonus No ads for 60 days $9.99
- Max 350 credits+40% bonus No ads for 180 days $24.99
Pay once, use for 12 months. Tax included. Never renews.
Ads stay off while you are signed in.
Sign in to buyWe email you tips and news about our tools. Unsubscribe anytime.
Compare the packsWe email you tips and news about our tools. Unsubscribe anytime.
Keep this tab open until it finishes.
We email you tips and news about our tools. Unsubscribe anytime.
Need more? Email contact@uncrop.si about pricing
The prompt
More text tools
How to turn an image into a prompt
Open the image
Choose a photo, drawing, render or screenshot, drop it on the box, or press Ctrl + V (Cmd + V on a Mac) to paste one.
Pick what to write
Detailed prompt is a full paragraph; Short prompt is one line. Keyword tags, Alt text and Description are there too.
Press Write the prompt
A smaller copy of the image goes to our AI, which describes what it sees and sends the words back.
Copy and use it
Copy the words, change any you like, and paste them into an image generator, a web page or a document.
What an image to prompt generator does
An image to prompt generator looks at a picture and writes the words an AI image generator would need to make something like it: the main subject, where it is, how the shot is framed, the light, the colors, the mood and the style, such as a photo, a watercolor or a 3D render. People use it to recreate a look they like, to start from a reference image instead of a blank box, to learn how good prompts are put together, and to keep a set of images in one style.
The same reading of the image gives four other useful results. Short prompt is a one-line version. Keyword tags is a comma-separated list for search, stock sites or tag-based generators. Alt text is one sentence for the alt attribute of an image on a web page. Description is a few plain sentences for someone who cannot see the picture.
Prompts, tags, alt text or a description: which to pick
Each style answers a different need:
| Style | What you get | Use it for |
|---|---|---|
| Detailed prompt | One paragraph of about 60 to 120 words | Recreating an image as closely as an image generator can |
| Short prompt | One line of about 15 to 30 words | A quick starting point you build on yourself |
| Keyword tags | 15 to 30 comma-separated words and phrases | Tag-based generators, stock photo keywords, your own image library |
| Alt text | One sentence of up to 125 characters | The alt attribute of an image on a web page or in an app |
| Description | 3 to 5 plain sentences | Captions, notes, and readers who cannot see the image |
What it can and cannot do
The words describe what can be seen. An image generator given them will make a new image in the same spirit, not a copy: the same prompt gives a different picture each time, and different generators read prompts differently. Expect to adjust a few words to get closer.
People are described only by what they wear, do and look like. The AI does not name anyone, does not guess who a person is, and does not guess age, ethnicity or other personal traits. Text in the image is quoted only when it can be read.
Everything is written in English. Small details in a busy picture can be missed, and the AI can be wrong about what something is, so read the result before you use it. Alt text in particular should say what matters on your page, which only you know.
Get a better prompt
Use a clear, uncropped image at a good size: before sending, your browser shrinks it to at most 1,600 pixels on its longest side, which keeps the main subject sharp. Crop away anything you do not want described with the crop tool, and blur faces you do not want sent with blur faces.
To find out whether an image you found was made with AI in the first place, check it with the AI image detector. To copy the words written in an image instead of describing it, use image to text.
Questions
How do I turn an image into a prompt?
Open the image on this page, keep Detailed prompt, and press Write the prompt. In a few seconds you get a paragraph that describes the subject, setting, lighting, colors and style, ready to copy into an AI image generator.
Is this image to prompt generator free?
Yes. Image to prompt shares the free AI image reads of the text tools: 10 a day without signing up, or 30 a day when you sign in with Google, which is free. Each image counts as one, and the count resets at midnight UTC.
Can it write alt text for an image?
Yes. Choose Alt text and you get one plain sentence of up to 125 characters that says what the image shows. Read it before you use it: good alt text also says why the image is on your page, which only you know.
Can it describe an image for me?
Yes. Choose Description for three to five plain sentences about what the image shows, where, who or what is in it and what they are doing, the colors and mood, and any text that can be read.
Will the prompt recreate my image exactly?
No. An image generator makes a new picture from the words, so you get something in the same spirit, not a copy, and each run is different. Adjust a few words to steer it closer.
Does it identify the people in a photo?
No. People are described only by what can be seen, such as their clothes and what they are doing. The AI does not name anyone or guess who they are, their age or their background.
Is my image uploaded or kept?
Only when you press Write the prompt. Your browser first makes a smaller copy of the image without its location or camera details, then sends it to the AI service we use, which describes it and sends the words back. We keep neither the image nor the words.
Why is there a daily limit?
Each read runs a paid AI model, so a daily limit keeps the tool free for everyone. Signing in with Google, which is free, raises it to 30 a day.