autobrr
autobrr redefines download automation for torrents and Usenet, drawing inspiration from tools like trackarr, autodl-irssi, and flexget. We've combined the best of these worlds into one versatile tool that can do it all, and then some.

We also have a light theme. Check it out here.
Full documentation can be found at https://autobrr.com
What is autobrr, and how does it fit into the ecosystem?
We can start by talking about torrent trackers (hereby referred to as indexers) and maintaining ratio. You are required to maintain a ratio with most indexers. Ratio is built by seeding your torrents. The earlier you're seeding a torrent, the more peers you make yourself available to on that torrent.
Software like Radarr and Sonarr utilizes RSS to look for new torrents. RSS feeds are updated regularly, but too slow to let you be a part of what we call the initial swarm of a torrent. This is where autobrr comes into play.
Many indexers announce new torrents on their IRC channels the second it is uploaded to the site. autobrr monitors such channels in real time and grabs the torrent file as soon as it's uploaded based on certain conditions (hereby referred to as filters) that you set up within autobrr. It then sends that torrent file to a download client of your choice via an action set within the filter. A download client can be anything from qBittorrent and Deluge, to Radarr and Sonarr, or a watch folder.
When your autobrr filter is set to send the torrent files to Radarr and Sonarr, they will decide if it's something they want, and then forward it to the torrent client they are set up with.
autobrr can also send matches (torrent files that meets your filter's criteria) directly to torrent clients like qBittorrent, Deluge, r(u)Torrent and Transmission. You don't need to use the *arr suite to make use of autobrr.
A lot of indexers do not announce new torrents in an IRC channel. You can still make use of these indexers with autobrr since it has built in support for feeds as well. Both Torznab, Newznab and regular RSS is supported. RSS indexers are treated the same way as regular indexers within autobrr.
Usenet support via Newznab feeds allows you to easily manage everything in a single application. While there is a lot of applications that handles RSS well, we think autobrr offers very easy to use filtering to help you get the content you want.
You can use Usenet feeds and send to arrs or send directly to SABnzbd.
For comprehensive installation instructions, visit our Installation Guide. This guide provides detailed steps for different platforms, including Windows, Linux, Docker, and more.
Remember to head over to our Configuration Guide to learn how to set up your indexers, IRC, and download clients after you're done installing.
Swizzin users can simply run:
sudo box install autobrr
Saltbox users can simply run:
sb install autobrr
For more info check the docs
QuickBox users can simply run:
qb install autobrr -u ${username}
For more info check the docs
box install autobrrbox install autobrrWe have support for a couple of other providers out of the box.
Please contact us on Discord if your provider is missing.
The scripts require some input, but do most of the work.
wget https://gobrr.sh/install_feral && bash install_feral
wget https://gobrr.sh/install_bytesized && bash install_bytesized
For other providers the Seedbox.io installer should work. If not, open an issue or contact us on Discord
wget https://gobrr.sh/install_sbio && bash install_sbio
Modify accordingly if running with unRAID or setting up with Portainer.
BASE_DOCKER_DATA_PATH to match your setup. Can be simply ./dataCreate docker-compose.yml and add the following. If you have an existing setup change to fit that.
version: "3.7"
services:
autobrr:
container_name: autobrr
image: ghcr.io/autobrr/autobrr:latest
restart: unless-stopped
environment:
- TZ=${TZ}
user: 1000:1000
volumes:
- ${BASE_DOCKER_DATA_PATH}/autobrr/config:/config
ports:
- 7474:7474
Then start with:
docker compose up -d
[!CAUTION] This image comes without a shell, and external filtering and actions relying on
execwill therefore not work with anything but compiled static binaries.To clarify:
BASHandSHshell scripts WILL NOT WORK!Use the standard image if you rely on this functionality.
For users who prioritize container security, one of the longterm maintainers offer alternative Docker images built on Distroless. Specifically the distroless/static-debian12:nonroot base image.
Distroless images do not contain a package manager or shell, thereby reducing the potential attack surface and making them a more secure option. These stripped-back images contain only the application and its runtime dependencies.
The repository for these builds can be found here: https://github.com/s0up4200/autobrr-distroless
Check the Windows Setup Guide here.
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
brew install autobrr
brew services start autobrr
Download the latest release, or download the source code and build
it yourself using make build.
wget $(curl -s https://api.github.com/repos/autobrr/autobrr/releases/latest | grep download | grep linux_x86_64 | cut -d\" -f4)
Run with root or sudo. If you do not have root, or are on a shared system, place the binaries somewhere in your home directory like ~/.bin.
tar -C /usr/local/bin -xzf autobrr*.tar.gz
This will extract both autobrr and autobrrctl to /usr/local/bin.
Note: If the command fails, prefix it with sudo and re-run again.
On Linux-based systems, it is recommended to run autobrr as a sort of service with auto-restarting capabilities, in order to account for potential downtime. The most common way is to do it via systemd.
You will need to create a service file in /etc/systemd/system/ called autobrr.service.
touch /etc/systemd/system/autobrr@.service
Then place the following content inside the file (e.g. via nano/vim/ed):
```systemd title="/etc/systemd/system/autobrr@.service" [Unit] Description=autobrr service for %i After=syslog.target network-online.target
[Service] Type=simple User=%i Group=%i ExecStart=/usr/bin/autobrr --config=/home/%i/.config/autobrr/
[Install] WantedBy=multi-user.target
Start the service. Enable will make it startup on reboot.
```bash
systemctl enable --now autobrr@$USER.service
By default, the configuration is set to listen on 127.0.0.1. While autobrr works fine as is exposed to the internet,
it is recommended to use a reverse proxy
like nginx, caddy
or traefik.
If you are not running a reverse proxy change host in the config.toml to 0.0.0.0.
The following environment variables can be used:
| Variable | Description | Default |
|---|---|---|
AUTOBRR__HOST |
Listen address | 127.0.0.1 |
AUTOBRR__PORT |
Listen port | 7474 |
AUTOBRR__BASE_URL |
Base URL for reverse proxy | / |
AUTOBRR__LOG_LEVEL |
Log level (DEBUG, INFO, WARN, ERROR) | INFO |
AUTOBRR__LOG_PATH |
Log file location | /config/logs |
AUTOBRR__LOG_MAX_SIZE |
Max size in MB before rotation | 10 |
AUTOBRR__LOG_MAX_BACKUPS |
Number of rotated logs to keep | 5 |
AUTOBRR__SESSION_SECRET |
Random string for session encryption | - |
AUTOBRR__CUSTOM_DEFINITIONS |
Path to custom indexer definitions | - |
AUTOBRR__CHECK_FOR_UPDATES |
Enable update checks | true |
AUTOBRR__DATABASE_TYPE |
Database type (sqlite/postgres) | sqlite |
AUTOBRR__DATABASE_DSN |
Database connection string. Use this or individual vars | - |
AUTOBRR__POSTGRES_HOST |
PostgreSQL host | - |
AUTOBRR__POSTGRES_PORT |
PostgreSQL port | 5432 |
AUTOBRR__POSTGRES_DATABASE |
Postgre |
$ claude mcp add autobrr \
-- python -m otcore.mcp_server <graph>