Get a caption from a photo, without writing a prompt

Upload the photo and CAPTD writes the caption from what is in it: the subject, the light, the setting. You don't write a prompt; you can add a line of notes if you want to. It's for anyone posting their own pictures, and the first 10 are free with no account.

A real example

Real CAPTD output, unedited

Only the photo, no notes at all, for Instagram. Everything below came from the picture.

Brooklyn Bridge walkway with steel cables, looking back at the Lower Manhattan skyline

What went in

Photo
Brooklyn Bridge walkway
Notes
none
Platform
Instagram
Tone
Story
Style
General

What came back

Hook

Looking up into that metal frame.

Caption

Some structures make you stop just to trace the geometry with your eyes. The scale of the metalwork against that wide open blue is hard to ignore.

Hashtags

#bridges #steelwork #urbanarchitecture #bridgephotography #archilovers

Alt text

  1. Short:A wide-angle view from a suspension bridge shows vertical steel cables stretching down to a roadway with a distant skyline across calm water under a clear blue sky.
  2. Medium:A wide-angle view from a high suspension bridge walkway shows thin vertical steel cables stretching down to a paved roadway below, with a dense urban skyline visible across calm water under a bright blue sky.
  3. Long:A high-angle view from a suspension bridge walkway shows thin vertical steel cables stretching down past a metal railing to a paved roadway below, with a dense urban skyline of tall buildings visible across calm water under a bright blue sky. Sloping diagonal cables cross the upper frame, dark informational panels line the foreground walkway edge, and distant stone bridge arches punctuate the horizon under an expansive clear blue sky.

Made in CAPTD's Snapshot on 8 November 2026, on a test build of the live app, and copied here as it came back.

How it works

  1. Drop the photo in. It's compressed in your browser, and all metadata, GPS included, is removed before upload.
  2. Leave the notes empty, or add the one thing the photo can't show: a name, a place, why it matters.
  3. Generate. If the caption isn't right, try another angle, regenerate just the caption or pick a different tone, and keep the rest.

Three ways to do it

Starting from a picture rather than from words.

Where the words come from

Writing it yourself
Your memory of the moment
A general AI chatbot
Your prompt, plus the photo if you attach it
CAPTD
The photo first; your notes, if any, on top

When there are no notes

Writing it yourself
Blank box
A general AI chatbot
Writes what the prompt asks for; you check it against the photo
CAPTD
Writes from what it can see; check any line about you that you didn't give it

Your photo's privacy

Writing it yourself
Stays with you
A general AI chatbot
Depends on the service
CAPTD
Metadata and GPS stripped in the browser before upload; uploads deleted within 90 days

FAQ

Do I have to type anything?
No. A photo on its own is enough. Notes help when the photo can't say something itself, like the name of a place or who is in it.
What happens to my photo?
It's resized in your browser and every bit of metadata, GPS included, is removed before upload. The copy CAPTD analyses never changes your original file.
Can it do several photos at once?
Yes: batch upload gives each photo its own caption, on paid plans (from 10 photos per batch on Basic). Slideshow writes one post with a caption for every slide.

Have the photo? Free for 10 posts on Instagram, TikTok and Facebook, no signup.