Upload an image, pick a style, and get back real, specific language — not a generic caption.
Click, drag and drop, or paste an image here
PNG, JPG, or WEBP up to 4MB
Select Description Style
Your description will appear here.
The AI actually looks at what's in the photo — it doesn't guess from the file name.
Subject, setting, lighting, color, texture, and mood all get read from the pixels themselves and written up in plain language.
Try it with your own photo →
A translucent glasswing butterfly with pale blue-tinted wings rests on a broad green leaf covered in fine water droplets, set against a dark, softly blurred rainforest background lit by a cool cyan glow.
Switch styles and get a genuinely different result from the same photo — brief, itemized, art-focused, or your own custom question.
See the description styles →
A lone figure stands on a canyon-side observatory balcony beneath a massive ringed planet at dusk.
Illustrative examples of the kind of output each style actually produces from one image.
A translucent glasswing butterfly with pale blue-tinted wings rests on a broad green leaf covered in fine water droplets, set against a dark, softly blurred rainforest background lit by a cool cyan glow.
A blue-tinted glasswing butterfly on a dew-covered leaf in a dark rainforest.
1. Butterfly (glasswing, blue-tinted wings) 2. Leaf (broad, green, water droplets) 3. Background foliage (blurred, dark) 4. Ambient light glow (cyan)
From upload to a usable description in under a minute.
Upload a file, drag it in, paste from your clipboard, or paste an image URL.
Detailed, brief, objects, art style, extracted text, or ask your own question.
It reads the actual visual content, not just a guess.
Drop the description into alt text, a caption, a catalog, or wherever you need it.
Straight answers about how this tool works.
No — for privacy reasons, "Describe the Person" deliberately avoids trying to identify anyone. It describes general appearance only (hair, clothing, expression, pose), never a name or identity.
It reads whatever text is clearly visible in the image, but isn't a dedicated OCR engine — very small, blurry, or heavily stylized text may be missed or misread.
Ask anything specific about the image — "What brand is this shoe?" or "What's happening in the background?" — and get a direct answer instead of a generic description.
PNG, JPG, or WEBP up to 4MB — click to browse, drag and drop, paste from your clipboard, or paste an image URL directly.
Yes — pick a language from the dropdown and the description is written in that language.
Yes — uploads with NSFW or otherwise disallowed content aren't supported.
