Repository navigation
faq
Is rDownloader open source? Yes — GNU GPL v3.0 or later. The source is at https://github.com/degoya/rDownloader.
Where can it run?
Windows x86-64, Linux x86-64 and ARM64, macOS Intel and Apple Silicon as native builds, and as a
Docker image for linux/amd64 and linux/arm64 — desktops, NAS devices, a Raspberry Pi with a
64-bit OS and servers alike. The minimum versions, and which combinations are tested, are on the
support matrix.
Does it include a web interface?
Yes, embedded in the server binary — nothing separate to deploy. It is a responsive single-page
app at http://127.0.0.1:8710 by default.
Which languages is the interface available in? English, German, French and Spanish. Auto-detected from the browser, switchable at any time.
Does it need an internet-facing server? No. It binds to loopback by default and is designed to be used locally. Remote access means putting a TLS reverse proxy in front — see Reverse proxy.
Can several people use it with separate accounts and roles? No. There is one administrator account; scripts, dashboards and assistants get scoped API tokens instead.
Which sources are supported?
Direct HTTP/HTTPS, one-click hosters and multihosters, Usenet/NZB, BitTorrent, media sites via
yt-dlp, image galleries via gallery-dl, livestream recordings via streamlink, and FTP, FTPS,
SFTP and WebDAV.
Do I need a premium hoster account? Mostly not. Fourteen of the bundled hoster resolvers, the XFileSharing sites among them, and MEGA can download through the anonymous flow — countdown and captcha included. An account lifts the speed and concurrency limits. Multihosters need one by design, and so do the cloud drives and cloud-storage services (Google Drive, OneDrive, Dropbox, Box, pCloud, Put.io, Seedr). See Hosters and accounts.
Why does only one free download run at a time per hoster? Hosters allow one free download per IP; a second attempt would break the first one's session and earn both an IP block. See Hosters and accounts.
Are media and gallery tools included?
The Docker image ships them. The native builds do not: install yt-dlp, ffmpeg, ffprobe,
gallery-dl and streamlink yourself, drop them in the vendor folder, or switch on
managed tools.
Does rDownloader provide Usenet access? No. It is a client — you bring your own NNTP provider and configure it under Settings → Usenet.
Can it monitor feeds and channels? Yes. Subscriptions poll channels, playlists, gallery profiles, RSS/Atom and podcast feeds, and Newznab/Torznab indexers, feeding hits into the LinkGrabber.
Why is nothing uploading on my torrents? Sharing with other peers is off by default — a new installation uploads nothing. Switch it on under Settings → BitTorrent. Downloading without ever uploading gets clients banned from some private trackers.
Can it connect to Sonarr, Radarr, Lidarr or Readarr? Yes. rDownloader answers as SABnzbd for Usenet and as qBittorrent for torrents. See Automation tool compatibility — including the one step people miss, creating the tool's category on this side.
What can an AI assistant control? Through the built-in MCP server, 265 tools — everything the interface can do, from the queue, the LinkGrabber and remote jobs to categories, rules, accounts, subscriptions, automations, plugins, the settings, the log and the diagnostic bundle's preview. What is left out, and why, is listed on the MCP server page.
Will an assistant see my passwords?
No. No MCP tool takes or returns a credential — a call carrying one is refused with
request.credential_rejected. You create the row over MCP and fill in the secret in the web
interface. You can also issue a read-only token, or one scoped to the queue without
api:secrets.
Can I give a script limited access? Yes. Tokens are minted with only the permission areas they need, and the issuing form opens on reading alone. Nothing confers stored credentials or administration implicitly.
Which plugin types are available? Twelve: resolvers, folder crawlers, OAuth providers, auth providers, notification destinations, post-processing steps, intake parsers, metadata enrichers, stream transforms, remote jobs, transfer backends and upload destinations. Seventy-two signed components ship with the release. See Plugin overview.
Is there an API?
Yes — REST under /api/v1 with a generated OpenAPI document, two Server-Sent Events streams, and
a command line client. See REST API.
Do I need a password, or can I use a passkey? Either. A passkey is an alternative to the password rather than an addition — the authenticator verifies a PIN or fingerprint, so one step carries both factors. The password keeps working alongside it, and an optional authenticator code (TOTP) is independent of both.
Can it run behind a reverse proxy, or under a sub-path? Yes, including the API, the event stream and the MCP endpoint. Tell it about the proxy under Settings → Security → Reachability & reverse proxy. See Reverse proxy.
Can I change the administrator password? Yes, under Settings → Security → Sign-in →
Administrator password, with the current one. The change ends every other session; API tokens keep
working. A forgotten password is reset on the computer rDownloader runs on, with rdownloader auth reset-password. See Accounts, sessions and
permissions.
Why do my passkeys not work at http://127.0.0.1:8710?
A relying party id must be a domain, and browsers reject an address. Use
http://localhost:8710, or set an external URL.
Can I put authentication at the proxy instead? Deliberately not supported. A proxy injecting an identity header would have to be trusted absolutely, and any request reaching the service directly would then be unauthenticated.
Where are my credentials stored?
In a local vault encrypted with XChaCha20-Poly1305; the master key comes from the OS keyring, with
a private 0600 fallback on Unix. Settings rows hold only references. See
Security and privacy.
Can I lock myself out with the login rate limiter? No. It counts failures per address, not per account, so nobody who reaches the login form can lock you out on purpose.
How do I upgrade? rDownloader tells you in the sidebar when a new version is out. The portable archive and the Windows installer install it themselves with Install and restart; with a package manager or the container image the dialog shows the command. See Updates.
What do I back up?
The data directory (/config in Docker): the database, the encrypted secret store, the installed
plugins, the password list and your scripts. Separately, the settings bundle under
Settings → Backup & restore → Configuration is the portable copy of your configuration.
Where do I find out what the service actually did? Logs in the sidebar: every log record the service kept, redacted before it was stored, filtered by level, component, code, correlation id or a search in the message. Audit log beside it answers the other question — who asked it to. See Diagnostics and maintenance.
How do I give somebody enough to help me, without giving them a credential? Create a diagnostic bundle on the Diagnostic bundle tab of the Logs page. You see an inventory of what it would hold before it is written, can untick anything the recipient should not get, and nothing is uploaded anywhere — the file is written on your own machine. It never contains download file names or contents, credential-bearing URLs, or secrets of any kind.
Why are my downloads gone after a docker rm?
A storage root at a path that is not mounted as a volume lives in the container's writable layer.
rDownloader badges such a root NOT PERSISTENT in the routing view. See
Docker and NAS.
Why do my schedules fire at the wrong time in Docker?
TZ defaults to UTC. Set it to your own zone.
Is there an in-app updater or a plugin repository? For plugins, yes: Settings → Plugins checks the official plugin repository — and any repository you add — for updates and offers them on its Catalogue tab after a review of publisher, key and permissions; per plugin they can also install automatically. See Installing and trust. The application checks for new versions too, and the portable archive and the Windows installer install them on one click; see Updates.
Can I scrape it with Prometheus?
Yes. /api/v1/metrics carries the queue, transfer and error figures behind a permission area of
its own, api:metrics, that reaches nothing else — so a scrape token in a configuration file
cannot read anything more. A Grafana dashboard for it comes with the repository; see
Statistics.
A settings page is not where I remember it. The settings are grouped into six rubrics — General, Downloads, Sources & protocols, Integrations, Network & security, Administration — and each page has an address of its own. The overview at Settings lists every page. The grouping is in Defaults and limits.
Something is not working and I do not know where to start.
Troubleshooting is organised by symptom. rdownloader doctor and
the readiness card on Settings → System → Status name most misconfigurations directly, and the
log viewer has the detail.
Getting started
- Installation
- Building from source
- Translating rDownloader
- First run
- Your first download
- The interface
Using rDownloader
- The download queue
- Download history
- The LinkGrabber
- Site rules
- Hosters and accounts
- Usenet and NZB
- BitTorrent
- Video and audio downloads
- Image galleries
- Livestream recording
- FTP, SFTP and WebDAV
- Object storage
- Subscriptions
- Remote jobs
- Storage, categories and rules
- Watched folders
- Post-processing
- Bandwidth and power
- Notifications
- Automation
- Statistics
- Keyboard and accessibility
Integrations
- Browser extension
- Desktop capture agent
- Command line client
- REST API
- MCP server
- Automation tool compatibility
Operating
- Docker and NAS
- Synology NAS
- Reverse proxy
- Accounts, sessions and permissions
- Sign in with an identity provider
- External tools
- Backup and restore
- Updates
- Security and privacy
- Diagnostics and maintenance
- Troubleshooting
- Defaults and limits
- Support matrix
Plugins
Plugin development
- Developing a plugin
- Resolver
- Transfer backend
- Intake parser
- Intake parser with mirrors
- Folder crawler
- Remote job
- Sign-in
- OAuth sign-in
- Metadata enricher
- Notification destination
- Post-processing step
- Upload destination
- Stream transform
- Plugin reference
- Packaging, signing and distribution
- Running a plugin repository
Reference