Upload a picture and our AI image describer writes a clear image description: a one-line caption, a short summary, or a detailed paragraph that walks through everything in it.
Drop an image, or browse, or paste Ctrl+V
PNG, JPG, WEBP up to 10 MB. Resized to 768px in your browser before sending.
5 of 5 free tries left today
Add an image and press Describe image.
Results usually arrive in a few seconds.
Most photos online have no description at all. That’s a problem for people who use screen readers, for search engines that can’t see images, and for anyone who needs to explain a picture in words. This describer reads the image and writes the description for you, in plain language and in the order a person would explain it: the overall scene first, then the details.
It’s also a quick way to let AI describe image details you might miss: small text, background objects, the direction of the light.
It’s different from our image to prompt generator. A prompt is written for an AI image tool, full of style and camera terms. A description is written for people: it says what’s there without telling anyone how to recreate it.
Caption mode sums a picture up in a single sentence, handy for photo libraries, social posts and file notes.
Try it
Description · Caption · Unedited output
A cyclist rides along a winding coastal road toward a glowing orange sunset over the ocean.
Detailed mode starts with the overall scene, then works through the objects, any text, the colors and the light, the way you’d explain a picture to someone who can’t see it.
Short mode sits in between: two or three sentences.
Try it
Description · Detailed · Unedited output
A cluttered, rustic woodworking workbench is illuminated by the warm, golden glow of an adjustable metal desk lamp. The scene is set in a cozy, dimly lit workshop with wooden walls and shelves. In the center of the workbench sits a newly constructed, light-colored wooden birdhouse with a circular entrance hole. Surrounding the birdhouse are various woodworking tools and materials: a wooden hammer, a red tape measure, a hand plane, several wooden scraps, and wood shavings scattered across the surface. Several glass jars filled with nails and wooden dowels stand behind the birdhouse. On the left, a potted green plant sits near a window, and a speckled ceramic mug rests on the bench. In the background, a tool rack hangs on the wall, displaying a collection of screwdrivers, pliers, and saws. The lighting is focused and dramatic, casting soft shadows and highlighting the textures of the wood and metal. The overall mood is industrious, nostalgic, and peaceful, suggesting a quiet moment of craftsmanship in a dedicated maker's space.
Upload, drag, paste or link a PNG, JPG or WEBP. Your browser shrinks it to 768px before sending.
Caption for one line, Short for a few sentences, Detailed for everything. Pick the language you want it written in.
A vision model reads the picture and writes the description, usually in two to five seconds.
Paste it into your caption, listing, report or notes. Edit anything you know better than the model.
Describe product photos, old family pictures or archive scans so they can be searched and catalogued.
Low vision, a tiny screenshot, or an unfamiliar scene: Detailed mode walks through everything in it.
Product listings, reports, archives and catalogs all need consistent descriptions. Start from a draft instead of a blank page.
Compare your own description with the model’s to spot details you missed and words you could use.
Use the image to prompt generator to get a prompt for Midjourney, Flux, Stable Diffusion or ChatGPT.
Yes. Add your picture to the describer above, choose a length, and ask it to describe this image. A vision model looks at the picture and writes what it sees: the subjects, what they’re doing, the setting, colors, any readable text and the overall mood. It takes a few seconds and needs no account.
Start with the big picture (what kind of image it is and the main subject), then move to the details in a logical order: people or objects, what’s happening, the setting, colors and light, and any text. End with the mood or purpose. The Detailed mode follows this order, so you can use its output as a model for your own writing.
Short gives one or two sentences that cover the main subject and setting. Detailed gives a full paragraph of roughly 120 to 200 words. Caption gives a single line that sums the picture up.
For the alt attribute, use our alt text generator instead: it gives three options under 125 characters, can use a focus keyword and page context, and scores each one. The describer is for fuller descriptions.
Accurate on the main subject, setting, colors and style in most images. The model can still miscount objects, misjudge small details, or misread stylized text, and it describes rather than identifies private people. Treat the result as a strong first draft.
It quotes short, clear text such as signs, labels and headings. It isn’t an OCR tool: long documents, handwriting and tiny print are often summarized or misread.
No. The image is resized in your browser to 768px, sent over an encrypted connection, described by an AI model on Cloudflare Workers AI, and not saved by us.
We count visits without cookies. With your OK we also use analytics cookies to see which pages and buttons people use. They never see your images, your text or the results. Details