Same-origin requests are supported. Private network addresses and non-HTTP protocols are blocked. Responses have time, size, and image-count limits. This API parses HTML; it does not run a headless browser or bypass access restrictions.
Find image URLs in public webpage HTML with /api/extract
Send a public HTTPS or HTTP URL. The endpoint discovers image tags, lazy-loaded attributes, responsive sources, inline CSS backgrounds and site icons, then returns a deduplicated list.
This example runs on a page served by imageextractor.top. Requests sent by a browser from another website origin are not supported. No API key is required, but request guards and processing limits still apply.
const response = await fetch("/api/extract", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ url: "https://example.com" })
});
const result = await response.json();
if (!response.ok) throw new Error(result.error);
console.log(result.images);Website image URL, format and dimension response fields
{
"source": "https://example.com/",
"images": [
{
"url": "https://example.com/photo.jpg",
"name": "photo.jpg",
"type": "jpeg",
"source": "https://example.com/",
"width": 1200,
"height": 800,
"size": 0
}
]
}Dimensions in this JSON come from HTML attributes and may be zero or represent display sizes. The tool workspace separately loads images to read natural dimensions and file sizes. A size of zero in the API response means the image has not been downloaded for analysis.
Download a discovered public image with /api/image
Fetch a public image through the same server. Use the optional download parameter to suggest a filename. Only supported image content is returned, with a bounded response size.
/api/image?url=https%3A%2F%2Fexample.com%2Fphoto.jpg
/api/image?url=https%3A%2F%2Fexample.com%2Fphoto.jpg&download=photo.jpgUnderstand website HTML discovery and local file processing
Invalid input, inaccessible sites, blocked destinations and oversized responses return an HTTP error with a readable message. Treat website content as untrusted and handle non-success responses. A website may require JavaScript, authentication or permissions that this endpoint cannot supply.
PDF and Office extraction, EXIF, palettes and OCR run in the browser; they have no server upload endpoint. Use the browser image tools for those tasks. The website endpoint discovers candidate URLs rather than guaranteeing that every remote image can be retrieved or reused.
Handle website image API errors and request limits
Check the HTTP status before using a response. An invalid URL or blocked private destination returns 400; a mismatched browser Origin returns 403; a non-JSON extraction request returns 415. A 429 response means the request or concurrency limit was reached, so pause before retrying instead of sending the same request repeatedly. Error JSON uses an error field with a readable message.
The current application guard allows up to 20 extraction requests per minute and 600 image requests per minute for a client key, with at most 30 requests in flight across the server process. Hosting or source websites can impose tighter limits. Process one public webpage URL per extraction request and keep the page's source URL alongside downloaded results.