Why download links expire: 403, 404 and CORS

Why does a video play but refuse to download?

A player can display a video through a media URL that is temporary, signed, or tied to the request that created it. Opening the page and downloading the file are therefore different operations, even when both show the same video.

The usual failures look similar but are not. A 403 is the source refusing a request; a 404 says the address was not found; CORS is a browser rule about which scripts may read a response. Knowing which layer failed tells you whether to retry the page, request the file again or stop.

Rule: a page that plays is not evidence that a copied media URL will work elsewhere, and ClapClip's own save links last 300 seconds. Treat the page URL as the stable input and the media URL as a one-time result.

Which links expire, and which ones do not?

Links expire when a source delivers media through signed or short-lived addresses, separate streams or server-side checks. ClapClip's own prepared files work the same way: the task is temporary and the save link is temporary.

Nothing here bypasses a private, login-only, region-locked, live or DRM-protected source. Copying a media URL is not a licence, and finding one in a browser's network panel is not proof that a complete file is available.

Rule: if the link required your session to play, expect it to fail outside that session, and remember that the form accepts only public HTTP(S) URLs up to 2,048 characters. Use the page URL and let a fresh analysis create a new selection.

How long do ClapClip's links last?

The form accepts a public HTTP(S) URL without embedded credentials, up to 2,048 characters. The prepared task is not permanent hosting, and both the task record and the save link have a clock.

• Task record: defaults to a one-hour expiry in the database. After that, the task and its result are no longer available.
• Save link: each file request redirects to a signed object URL valid for 300 seconds.
• Selection: if the analysis selection expires, the site asks you to find the video again instead of reusing the old choice.

Two different clocks are easy to confuse. The task record holds the result; the save link is a short-lived doorway to the stored object. When the doorway expires, the result can still be there; when the container expires, no new doorway can be opened.

Rule: if a save link is more than 300 seconds old, request the file again from the task page while the task is alive. If the task itself has expired, start from the page URL and run a new analysis.

403, 404 or CORS: which error are you seeing?

The three errors point at different layers, and the right next step depends on which one you got. The status code is more useful than the browser's generic failure message.

• 403 Forbidden: the source refused the request. The page may still exist; the server was not allowed to fetch it.
• 404 Not Found: the address was not found. This is often an expired media URL rather than a deleted page.
• CORS: a script on one origin could not read another origin's response. Moving the request to a server avoids the browser rule, but the source can still refuse that server.

Rule: 403 and sign-in walls are access problems; 404 on a media URL is usually a freshness problem; CORS is a browser problem. None of them is fixed by starting duplicate tasks.

How do you recover from an expired link?

Confirm that you still have the original public page URL, not a player fragment, a shortened redirect or an old media address. Do not strip query parameters or add credentials.

Run a fresh analysis and choose a format the tool actually returns. If the page opens but analysis keeps failing, record the platform and the error instead of retrying blindly.

Save the prepared file promptly. The save link lasts 300 seconds and the task record itself expires, so a ready message alone does not put the file on your device.

If the source now requires login, is private, is region-restricted or is gone, stop. An anonymous tool cannot restore access, and support should never receive cookies, passwords or tokens.

Rule: retry the page, not the media URL. A new analysis is the only retry that can produce a fresh 300-second save link.

What should you keep instead of a media URL?

Keep the page URL. It is the durable identifier for the video and the input a new analysis needs. A discovered media address is the output of one inspection and may stop working without the page changing.

Rule: a save link older than 300 seconds is expired; re-request it from the task page or re-analyse the page. The trade-off is that the online service cannot promise permanence or universal access. It offers a bounded, anonymous path to the first 300 seconds at up to 1280×720, and it fails closed when a source refuses.

Bottom line: archive the page link, not the stream link. If you need a file to survive longer than the task, save it to your own device before the 300-second window closes.