Caption an image
JPEG, PNG, GIF or WebP. Large photos are shrunk in your browser before upload, so nothing oversized leaves your device.
The four styles
Same image, four different jobs. Pick the one that matches where the text is going.
| Style | What you get | Use it for |
|---|---|---|
| Alt text | One sentence under 125 characters, describing what matters visually. | The alt attribute on a web image — screen readers and SEO. |
| Social caption | One or two engaging sentences, plus 3–5 hashtags. | Instagram, Facebook, LinkedIn posts. |
| Ad copy | A headline, a benefit-led body line, a call to action, and hashtags. | Paid social and display creative. |
| Plain description | Two or three factual sentences with no marketing language. | Cataloguing, image libraries, internal notes. |
Frequently asked questions
How the caption generator works.
What happens to the images I upload?
Your image is sent to the OpenAI API to be read, and the caption comes straight back to your browser. It is not saved to this website, and no copy is kept on the server after the caption is generated.
Why is my photo resized before uploading?
Photos straight off a phone are often far larger than the model needs. Your browser scales anything over 1568 pixels on its long edge down to that size first, which makes the upload faster and the request cheaper without any loss in caption quality.
Is there a usage limit?
Yes. Each visitor can generate a limited number of captions per hour. Every caption costs the site owner money in API usage, so the cap keeps that predictable.
Can I edit the caption?
Yes — the result box is editable. Tweak the wording, then use Copy. Treat every caption as a first draft and check it against the image before publishing.
Which file types work?
JPEG, PNG, GIF and WebP — the four formats the vision model accepts. The file’s real contents are checked on the server, not just its extension.