Repository navigation
first download
This page walks one link from pasting it to the finished file on disk. It assumes you have completed the setup assistant and have at least one storage root. There are two ways in.
On the Downloads page click Direct job at the top, or press a. Paste a direct HTTP(S) link or a magnet link into the dialog and click Enqueue. The transfer starts immediately, in the default category.
Use this for a link you already trust and do not need to review — a plain file URL, a magnet.
Everything else goes through the LinkGrabber first. It is the review stage: links are grouped, checked and given a category before a single byte is fetched.
-
Open LinkGrabber (sidebar, or press 2).
-
Click Add links. Paste one or more links, or any text with links embedded in it — rDownloader picks them out. Multi-part archives (
file.part1.rar,file.part2.rar, …) are grouped into one package automatically.- Package name (optional) — leave it empty and the name is derived from the file names.
- Archive password (optional) — set it now and unpacking will use it later.
-
Confirm with Add to LinkGrabber.
-
The links appear as a package. An online check runs automatically; each link ends up as online, offline, already in the list, not checked or not resolvable. You can re-run it with Check links. Only not resolvable cannot be queued — that address answers with a web page, not with a file.
If the links came from a release page, several of them may be the same file at different hosters. Those arrive as one row with the rest behind it as mirrors; see The LinkGrabber.
-
Set the package's Category and Priority if you want something other than the default. The category decides where the files land.
-
Click Enqueue on the package — or Enqueue all for everything in the list. Use Add paused instead if you want the entries in the queue without starting them.
A link with no matching account shows No account – will try a free/direct download. That is normal for the bundled hosters; see Hosters and accounts.
The package is now on the Downloads page (1). Each file is a row with its state, progress, speed and estimated time remaining. Above them:
- Live — the current combined speed of everything running.
- Limit — the global speed limit in MiB/s. Leave it empty for unlimited; it takes effect immediately, no restart.
- Start all / Pause all — the global control. p does the same from anywhere.
- Filters — All, Active, Waiting, Completed.
Each package has its own start/pause buttons, and so does each file. A package shows
{finished} of {total} files finished, its own speed and its own estimate.
Rows move through a defined set of states — Queued, Downloading, Verifying, Extracting, Completed and so on. Transfer states lists all fourteen and what moves a row between them.
<storage root>/<category path>/<package name>/<file name>
Every package gets its own directory. Click the destination on the package header to copy the full path to the clipboard.
If the package needs post-processing — a PAR2 repair, an unpack, a checksum check — that happens after the last file arrives and is shown as its own progress on the package. See Post-processing.
- A link is offline. The online check said so before you queued it. Delete it, or look for a mirror.
-
A transfer failed. It is retried automatically — eight times by default, honouring the
server's
Retry-After. After that the row sits in Failed with the reason. Where the package holds mirrors of that file, the next one takes over by itself rather than the download stopping. - A hoster's limit was reached. A daily or rate limit and an IP block are waited out — an hour when the hoster names no reset, a quarter of an hour for an IP block — and the wait costs none of the retries; the row shows Next attempt at with the time. Only after 48 waits in a row does it give up.
-
You want failed downloads to come back by themselves. Switch on Settings → General →
Retry failed downloads automatically. A download that failed on a limit, an IP block, an
unreachable server or a file reported offline goes back into the queue after the interval you
set (1–24 hours, 6 by default), for a number of rounds (3 by default,
0for no limit), each with a fresh retry budget. A deleted link, a refused account, a wrong checksum or a rejected captcha never comes back on its own. The row shows when the next round is due. - A captcha is waiting. An image or click captcha is answered in this interface. A widget — reCAPTCHA, hCaptcha, Turnstile — is answered in your own browser through the browser extension, or by a solver service. See Hosters and accounts.
- You want to start over. Use Reset on the row or the selection. Partial data, checkpoints, progress, the retry budget and the package's post-processing steps are discarded and the transfer begins from zero. A finished file is kept unless the confirmation explicitly asks for it to be deleted.
- Nothing starts at all. Check the readiness card under Settings → System → Status: a full storage root, a switched-off service or a missing external tool is named there. See Troubleshooting.
Once this works by hand, automate the intake:
| Route | Page |
|---|---|
Click'n'Load, clipboard, .nzb double-click |
Desktop capture agent |
| A context-menu entry in the browser, intercepted downloads | Browser extension |
A folder you drop .nzb, .torrent or container files into |
Watched folders |
| Channels, playlists, feeds, indexers and release pages, polled on a schedule | Subscriptions |
| A script or another machine | Command line client, REST API |