Free AI Image Describer
Drop in a photo and get a clear written description, generated by the AI built into your browser. Your images never leave this device.
0 images sent anywhereEverything runs on this device. Your images never leave this computer.
Needs desktop Chrome (Windows, macOS, Linux or ChromeOS). On a phone or another browser, open this page on your computer. Chrome downloads its AI model once, which varies in size and needs about 22 GB of free disk space.
How it works
Add one image or a whole batch. The tool reads each file locally in your browser and passes it to the AI that ships inside Chrome. There is no upload step because there is no server that receives your pictures.
For each image it writes a description of who or what is in the frame, where things are, what is happening and which text is visible. You can ask for a short two-sentence summary or a more detailed paragraph, and switch the output language.
Descriptions are drafts. Click any result to edit it, copy it, or add alt text and hashtags from the same pass if you also need them. Nothing is saved anywhere except in your own browser.
The first time you use it, Chrome downloads its on-device AI model. The download size varies, and Chrome needs about 22 GB of free disk space to install it. The model is shared by every site that uses it, so you only set it up once. It needs desktop Google Chrome (Windows, macOS, Linux or ChromeOS) with its built-in AI available. Phones and other browsers are not supported yet.
What the image describer does
- Writes a full description of each image: subject, setting, actions, colours and any readable text.
- Short or detailed length, so you can get a caption-sized summary or a paragraph.
- Optional alt text and hashtags from the same run if you want all three.
- Output in English, Spanish or Japanese.
- Add a line of page context to steer the wording toward what matters for your page.
Example results
Real examples of what this preset writes. Results are drafts: read them and edit before you publish.
harbour-morning.jpg
Example photo: A photograph of small fishing boats moored in a calm harbour at sunrise, with hills in the background.
DescriptionSeveral small wooden fishing boats rest on still water in a harbour. Low golden light from the left catches their hulls and reflects on the surface. Green hills and a few white houses sit behind them under a pale orange sky.Detailed length: subject first, then light, setting and background.
quarterly-dashboard.png
Example photo: A screenshot of an analytics dashboard with a line chart, three summary numbers and a navigation sidebar.
DescriptionA screenshot of a web analytics dashboard. A dark sidebar on the left lists menu items. The main area shows three summary tiles above a line chart that climbs steadily from left to right, with a legend underneath.It describes layout and structure; check any numbers it reads against the source.
market-stall.webp
Example photo: A woman arranging bunches of tulips on a wooden market stall, with a handwritten price sign.
DescriptionA woman in a green apron arranges bunches of pink and yellow tulips in buckets on a wooden market stall. A chalkboard sign next to her reads 'Tulips 5 per bunch'.Short length: two sentences, including the text on the sign.
Frequently asked questions
How is this different from other image describers?
Most online image describers upload your file to a server and put a limit on how many you can run. This one runs on the AI model inside your own browser, so there is no upload and no daily cap. The trade-off is that it needs desktop Chrome and a one-time model setup.
Can it describe screenshots, diagrams or photos with text?
Yes, it will mention visible text and the general layout of a screenshot or diagram. It is a small on-device model, so read the draft and fix details such as exact numbers before you publish.
How detailed can the descriptions be?
Choose short for one or two sentences, or detailed for a fuller paragraph covering subject, setting, mood and notable objects. If you add page context, the description leans toward what matters for that page.
Does it work offline?
After Chrome has downloaded its AI model, describing images does not need an internet connection. Loading this page does, so open it once while online.
Is anything stored or sent anywhere?
Your images never leave this device. Results are kept only in your own browser storage so you can come back to them, and you can clear them at any time. Ads on this page load outside the tool and never see your images.
Why does it say I need Chrome on a computer?
The AI is built into desktop Chrome. Other browsers and phones do not expose it yet, so the tool cannot run there. The examples below show what you would get.