Options

cross-seed has several program options, which can either be specified on the command line or in a configuration file. The priority is shown below.

CLI > Config File > Fallback

If you specify an option both on the command line and in the config file, the command line value will override the config file's value.

danger

DO NOT remove options from the config file, this will cause errors.

Options on the command line

All options on the command line have a long form, which looks like --long-form-option <value>. Some more common options have a single-letter short form as well, which look like -f <value>.

Some options take multiple arguments, which are always separated by spaces:

--variadic-option arg1 arg2 arg3

tip

Any references to options you see in camelCase have their corresponding long-form option name as the kebab-cased variant. For example, the excludeRecentSearch option's long-form CLI option name is --exclude-recent-search.

Options in the config file

For instructions on creating and editing the config file, use the getting started guide. DO NOT remove options from the config file, this will cause errors.

tip

What should I do after updating my config?

All options

`delay`*

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`delay`	`-d <value>`	`--delay <value>`	`number` (seconds)	`30`	`30`

When running a search with cross-seed search or using searchCadence in daemon mode, the delay option lets you set how long you want cross-seed to sleep in between searching for each torrent.

If you set it higher, it will smooth out the load on your indexer, however setting this lower will result in cross-seed running faster.

DISCLAIMER

You need to read your tracker's rules and be aware of their limits.

You and you alone are responsible for following your tracker's rules.

`delay` Examples (CLI)

cross-seed search -d 60
cross-seed search --delay 30

`delay` Examples (Config file)

delay: 30,

`torznab`*

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`torznab`	`-T <urls...>`	`--torznab <urls...>`	`string[]`	`[]`	`[]`

List of Torznab URLs. You can use Jackett, Prowlarr, or indexer built-in Torznab implementations.

This entry MUST be wrapped in []'s and strings inside wrapped with quotes and separated by commas.

caution

http://localhost:9117/p/a/t/h?query=string
└────────host───────┘└─path─┘└───query───┘

The path of each URL should end in /api.

Finding your Torznab URLs

For Prowlarr and Jackett you can simply copy the RSS URL from the WebUI.

note

This works because in Torznab, "RSS feeds" are just a search for the first page of unfiltered (no search query) results from the indexer.

`torznab` Examples (CLI)

cross-seed search --torznab https://localhost/prowlarr/1/api?apikey=12345
cross-seed search -T http://prowlarr:9696/1/api?apikey=12345 http://prowlarr:9696/2/api?apikey=12345
cross-seed search -T http://jackett:9117/api/v2.0/indexers/oink/results/torznab/api?apikey=12345

`torznab` Examples (Config file)

torznab: ["https://localhost/prowlarr/1/api?apikey=12345"],

torznab: [
    "http://prowlarr:9696/1/api?apikey=12345",
    "http://prowlarr:9696/2/api?apikey=12345",
],

torznab: ["http://jackett:9117/api/v2.0/indexers/oink/results/torznab/api?apikey=12345"],

`sonarr`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`sonarr`		`--sonarr <url(s)>`	`string[]`	`[]`	`[]`

NOTICE

Read about the functionality in the v6 Migration Guide

The URL to your Sonarr instance with your ?apikey= parameter appended to the end.

Finding your Sonarr URL

For Sonarr you can simply append ?apikey= to the end of your WebUI base URL with your API key after the =.

`sonarr` Examples (CLI)

cross-seed search --sonarr https://localhost/?apikey=12345
cross-seed search --sonarr https://localhost/?apikey=12345 https://localhost4k/?apikey=12345

`sonarr` Examples (Config file)

sonarr: ["https://sonarr/?apikey=12345"],

sonarr: ["http://sonarr:8989/?apikey=12345","http://sonarr4k:8990/?apikey=12345"],

`radarr`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`radarr`		`--radarr <url(s)>`	`string[]`	`[]`	`[]`

NOTICE

Read about the functionality in the v6 Migration Guide

The URL to your Radarr instance with your ?apikey= parameter appended to the end.

Finding your Radarr URL

For Radarr you can simply append ?apikey= to the end of your WebUI base URL with your API key after the =.

`radarr` Examples (CLI)

cross-seed search --radarr https://localhost/?apikey=12345
cross-seed search --radarr https://localhost/?apikey=12345 https://localhost4k/?apikey=12345

`radarr` Examples (Config file)

radarr: ["https://radarr/?apikey=12345"],

radarr: ["http://radarr:7878/?apikey=12345","https://radarr4k:7879/?apikey=12345"],

`useClientTorrents`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`useClientTorrents`	N/A	`--use-client-torrents`	`boolean`	`true`	`false`

warning

Deluge with Docker can have a desyncing issue with the WebUI daemon after prolonged uptime with useClientTorrents, if you notice strange behavior you may need to use torrentDir instead.

Query the torrent client APIs to find matches against its contents. This is a replacement for torrentDir and is generally recommended. This method supports qBittorrent's SQLite mode and all forms of content layouts/renames/modifications to the torrent in client.

`useClientTorrents` Examples (CLI)

cross-seed search --use-client-torrents
cross-seed search --no-use-client-torrents

`useClientTorrents` Examples (Config file)

useClientTorrents: true,

`torrentDir`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`torrentDir`	`-i <dir>`	`--torrent-dir <dir>`	`string`	`null`	`null`

qBittorrent

You MUST use useClientTorrents instead if any apply:

You are using SQLite in Preferences -> Advanced
You have renamed torrents or its files in client
You have torrents added WITHOUT Content Layout set to Original

Point this at a directory containing torrent files, or if you're using a torrent client integration and injection - your torrent client's .torrent file store/cache.

What directory is supported

More details on why using paths other than the .torrent store from the client's config/appdata directory (ex: 'state' or 'BT_backup') is not supported and discouraged can be found here.

If you don't know where your torrent client stores its files, please use the table below to find the client's store of .torrent files. This folder can then either be mounted as something like /torrents (with read only (:ro) permissions) as a "shorthanded" path if you're using Docker or used outright as your torrentDir.

Data-Only Searching

If you wish to search only your data, we previously recommended pointing this to an empty directory. You can now set this to null if you wish to search only your dataDirs.

Client	Linux	Windows	Mac
rTorrent	your session directory as configured in .rtorrent.rc	your session directory as configured in .rtorrent.rc	your session directory as configured in .rtorrent.rc
Deluge	`/home/<username>/.config/deluge/state`	%APPDATA%\deluge\state	current version of Deluge not officially supported
Transmission	`/home/<username>/.config/transmission/torrents`	Unknown (please submit a PR!)	Unknown (please submit a PR!)
qBittorrent	`/home/<username>/.local/share/data/qBittorrent/BT_backup`	`C:\Users\<username>\AppData\Local\qBittorrent\BT_backup`	`~/Library/Application Support/qBittorrent/BT_backup`

`torrentDir` Examples (CLI)

cross-seed search --torrent-dir ~/.config/deluge/state
cross-seed search -i ~/.config/transmission/torrents

`torrentDir` Examples (Config file)

torrentDir: "/home/<username>/.config/deluge/state",

torrentDir: "C:\\torrents",

WINDOWS USERS

It is necessary to insert double-slashes for your paths, back-slashes are "escape characters" and "\\" equates to "\".

`outputDir`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`outputDir`	`-s <dir>`	`--output-dir <dir>`	`string`	`null`	`null`

tip

Keep this set to outputDir: null for the best experience with cross-seed. This will map outputDir into your cross-seed config directory.

danger

DO NOT USE THIS DIRECTORY AS A WATCH FOLDER FOR YOUR TORRENT CLIENT!

With action: "inject", cross-seed will use this directory to retry injections only, it will be empty nearly all the time. With action: "save", cross-seed will store the .torrent files it finds in this directory. Only change this from null if you are using action: "save" and the path would be more convenient.

`outputDir` Examples (CLI)

cross-seed search
cross-seed search --output-dir /path/to/folder

`outputDir` Examples (Config file)

outputDir: null,

outputDir: "/path/to/folder",

WINDOWS USERS

It is necessary to insert double-slashes for your paths, back-slashes are "escape characters" and "\\" equates to "\".

`dataDirs`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`dataDirs`	`--data-dirs <dirs...>`	`--data-dirs <dirs...>`	`string(s)`	`[]`	`[]`

tip

Usages and Configuration

cross-seed will search the paths provided (separated by spaces). If you use Injection cross-seed will use the specified linkType to create a link to the original file in the linkDirs.

Docker

You will need to mount the volume for cross-seed to have access to the data and linkDirs.

`dataDirs` Examples (CLI)

cross-seed search --data-dirs /data/usenet/movies

`dataDirs` Examples (Config file)

dataDirs: ["/data/usenet/movies"],

dataDirs: ["/data/usenet/movies", "/data/torrents/tv"],

dataDirs: ["C:\\My Data\\Downloads\\Movies"],

WINDOWS USERS

It is necessary to insert double-slashes for your paths, back-slashes are "escape characters" and "\\" equates to "\".

`maxDataDepth`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`maxDataDepth`	N/A	`--max-data-depth <value>`	`number`	`2`	`2`

tip

What value should I set my maxDataDepth to?

Determines how deep into the specified dataDirs cross-seed will go to generate searchees. Setting this to higher values will result in more searchees and more API hits to your indexers.

`maxDataDepth` Examples (CLI)

cross-seed search --max-data-depth 2

`maxDataDepth` Examples (Config file)

maxDataDepth: 2,

`linkCategory`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`linkCategory`	N/A	`--link-category <category>`	`string`	`cross-seed-link`	`cross-seed-link`

cross-seed will use this category for all injected torrents when linking is enabled.

Docker

You will need to mount the volume for cross-seed to have access to the data and linkDirs.

`linkCategory` Examples (CLI)

cross-seed search --link-category category

`linkCategory` Examples (Config file)

linkCategory: "Category1",

`duplicateCategories`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`duplicateCategories`	N/A	`--duplicate-categories`	`boolean`	`false`	`false`

cross-seed will inject using the original category, appending '.cross-seed', with the same save paths as your normal categories. For qBittorrent with linking enabled, this will be applied as tag instead while keeping linkCategory.

info

This will prevent an Arr from seeing duplicate torrents in Activity, and attempting to import cross-seeds. You do not need to set this to true if you are using linking.

Example: if you have a category called "Movies", this will automatically inject cross-seeds to "Movies.cross-seed"

`duplicateCategories` Examples (CLI)

cross-seed search --duplicate-categories

`duplicateCategories` Examples (Config file)

duplicateCategories: true,

duplicateCategories: false,

`linkDirs`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`linkDirs`	N/A	`--link-dirs <dirs...>`	`string(s)`	`[]`	`[]`

tip

Ideally, you should only have a single linkDir and use drive pooling. Using multiple linkDirs should be reserved for setups with cache/temp drives or where drive pooling is impossible.

How to configure linking?

cross-seed will link (symlink/hardlink) in the path provided. If you use Injection cross-seed will use the specified linkType to create a link to the original file in one of the linkDirs during searches where the original data is accessible (both torrent and data-based matches). The correct linkDir is chosen by matching the stat.st_dev for the source and link paths. If no linkDir shares a stat.st_dev with the source, the injection will fail for hardlinks and fallback to the first linkDir for symlinks.

Docker

You will need to mount the volume for cross-seed to have access to the dataDir and linkDirs.

`linkDirs` Examples (CLI)

cross-seed search --linkDirs /data/torrents/xseeds /data1/torrents/cross-seed-links

`linkDirs` Examples (Config file)

linkDirs: ["/data/torrents/SomeLinkDirName"],

linkDirs: ["C:\\cross-seed-links", "D:\\xseeds"],

WINDOWS USERS

It is necessary to insert double-slashes for your paths, back-slashes are "escape characters" and "\\" equates to "\".

`linkType`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`linkType`	`--link-type <type>`	`--link-type <type>`	`string`	`hardlink`	`symlink`

tip

What linkType should I use?

cross-seed will link (symlink/hardlink/reflink) in the method provided. If you use Injection cross-seed will use the specified linkType to create a link to the original file in one of the linkDirs.

Valid methods for linkType are symlink, hardlink, and reflink.

tip

It's possible to use an alternative reflink method detailed here

`linkType` Examples (CLI)

cross-seed search --linkType hardlink

`linkType` Examples (Config file)

linkType: "hardlink",

linkType: "symlink",

`matchMode`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`matchMode`	`--match-mode <mode>`	`--match-mode <mode>`	`strict`/`flexible`/`partial*`	`flexible`	`strict`

cross-seed uses three types of matching algorithms strict, flexible, and partial. All types are equally safe. strict is required if you cannot use linking. Using partial will find all possible cross seeds of your data and allows setting seasonFromEpisodes to a value below 1.

note

These algorithms can only be ran if cross-seed has snatched the torrent files. The vast majority of candidates get rejected before a snatch has happened by parsing information from the title. Using partial minimizes the amount of wasted snatches since it's the most complete.

option	description
`strict`	requires all file names to match exactly along with file sizes
`flexible`	allows for file renames and slight inconsistencies
`partial`	can be read about in detail here

For media library searches flexible or partial is necessary due to the renaming of files.

`matchMode` Examples (CLI)

cross-seed search --match-mode flexible
cross-seed search --match-mode strict

`matchMode` Examples (Config file)

matchMode: "partial",

matchMode: "strict",

`skipRecheck`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`skipRecheck`	`N/A`	`--skip-recheck`	`boolean`	`true`	`true`

Set this to false to recheck all torrents upon injection. Set this to true to only recheck necessary injections such those from partial.

tip

Torrents will be resumed even with skipRecheck: false, if applicable.

`skipRecheck` Examples (CLI)

cross-seed search # will skip rechecking
cross-seed search --no-skip-recheck # will not skip rechecking

`skipRecheck` Examples (Config file)

skipRecheck: true,

skipRecheck: false,

`includeSingleEpisodes`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`includeSingleEpisodes`	`N/A`	`--include-single-episodes`	`boolean`	`false`	`false`

tip

It's recommended to use includeSingleEpisodes: false in your config and override it for webhook commands triggered on download completion. Combined with matching single episodes from announce, this should match all episodes without the downside of searching for trumped/dead torrents.

Set this to true to include ALL SINGLE episodes when searching (which are ignored by default).

info

This will NOT include episodes present inside season packs (data-based searches).

Behavior of this option has changed in v6, please see the migration guide for details on the implementation's changes'.

`includeSingleEpisodes` Examples (CLI)

cross-seed search --include-single-episodes # will include single episodes not from season pack
cross-seed search # will not include episodes

`includeSingleEpisodes` Examples (Config file)

includeSingleEpisodes: true,

includeSingleEpisodes: false,

`seasonFromEpisodes`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`seasonFromEpisodes`	`N/A`	`--season-from-episodes`	`number` (decimal from 0 to 1)	`1`	`null`

cross-seed will also aggregate individual episodes into season packs for searching (when applicable) or to match with season packs from rss/announce. This will only match season packs where you have at least the specified ratio of episodes. null disables this feature. If enabled, values below 1 requires matchMode partial.

tip

This feature works best with matchMode partial and Sonarr. You can avoid downloading the same missing episodes on multiple trackers by following these steps.

`seasonFromEpisodes` Examples (CLI)

cross-seed search --season-from-episodes 0.8 # will also combine episodes into season packs if you have at least 80%
cross-seed search --no-season-from-episodes # will not attempt to join episodes to season packs

`seasonFromEpisodes` Examples (Config file)

seasonFromEpisodes: 0.8, // requires 80% of the episodes to cross-seed a season pack

seasonFromEpisodes: null, // will disable season pack from episodes

`autoResumeMaxDownload`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`autoResumeMaxDownload`	`N/A`	`--auto-resume-max-download`	`number` (0 to 52428800)	`52428800`	`52428800`

The amount remaining for an injected torrent in bytes for cross-seed to resume. For torrents with a larger amount remaining, you will need to manually resume as you can avoid downloading the same missing data on multiple trackers by following these steps.

`autoResumeMaxDownload` Examples (CLI)

cross-seed search --auto-resume-max-download 0 # only resume complete matches

`autoResumeMaxDownload` Examples (Config file)

autoResumeMaxDownload: 52428800,

autoResumeMaxDownload: 0,

`ignoreNonRelevantFilesToResume`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`ignoreNonRelevantFilesToResume`	`N/A`	`--ignore-non-relevant-files-to-resume`	`boolean`	`false`	`false`

If set to true and the amount remaining is above autoResumeMaxDownload, resume if accumulated size of the missing files is less than the accumulated size of the known-irrelevant files such as nfo, sample, txt, subs, proofs, and bonus/commentary files present in the torrent.

note

The goal is to resume all torrents that meet this criteria. However, for safety, a 200 MiB limit has been set on the remaining download size. We're open to increasing this limit if needed.

`ignoreNonRelevantFilesToResume` Examples (CLI)

cross-seed search --ignore-non-relevant-files-to-resume # will resume if the only files missing are known-irrelevant files
cross-seed search --no-ignore-non-relevant-files-to-resume # will not resume if the only files missing are known-irrelevant files

`ignoreNonRelevantFilesToResume` Examples (Config file)

ignoreNonRelevantFilesToResume: true,
ignoreNonRelevantFilesToResume: false,

`includeNonVideos`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`includeNonVideos`	N/A	`--include-non-videos`	`boolean`	`false`	`false`

NOTICE

Behavior of this option has changed in v6, please see the migration guide for details on the implementation's changes.

Set this to true to include torrents that contain a majority of files other than video files (.mp4, .avi, .mkv) in the search.

`includeNonVideos` Examples (CLI)

cross-seed search --include-non-videos # will include non-videos
cross-seed search --no-include-non-videos # will not include non-videos
cross-seed search # will not include non-videos

`includeNonVideos` Examples (Config file)

includeNonVideos: true,

includeNonVideos: false,

`fuzzySizeThreshold`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`fuzzySizeThreshold`	N/A	`--fuzzy-size-threshold <value>`	`number` (decimal from 0 to 0.1)	`0.02`	`0.02`

Increase this number to reject fewer torrents based on size. There is no guarantee that it will increase your match rate.

`fuzzySizeThreshold` Examples (CLI)

cross-seed search --fuzzy-size-threshold 0.02
cross-seed daemon --fuzzy-size-threshold 0.02

`fuzzySizeThreshold` Examples (Config file)

fuzzySizeThreshold: 0.02,

`excludeOlder`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`excludeOlder`	`-x <value>`	`--exclude-older <value>`	`string` in the ms format	`2 weeks`	`undefined`

tip

How to ignore excludeOlder for a single search?

What's the best way to add new trackers?

If you prefer to spread out repeat searches for a longer safety net, use excludeRecentSearch: "90 days" and excludeOlder: "450 days".

When running a search, this option excludes anything first searched more than this long ago. This option is only relevant in search mode or in daemon mode with searchCadence turned on.

Note

Search history is stored on a per-indexer basis.

Searches that failed on specific indexers (for example - due to timeout or rate-limiting) will not be marked as having been searched, and thus will not be excluded by this setting for those specific indexers on the next run.

`excludeOlder` Examples (CLI)

cross-seed search -x 10h # only search for torrents whose first search was less than 10 hours ago or never
cross-seed search --exclude-older "3 days" # only search for torrents whose first search was less than 3 days ago or never
cross-seed search -x 0s # only search for each torrent once ever
cross-seed search --no-exclude-older # disables/overrides the excludeOlder value in config.js

`excludeOlder` Examples (Config file)

excludeOlder: "10 hours",

excludeOlder: "3days",

excludeOlder: "0s",

`excludeRecentSearch`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`excludeRecentSearch`	`-r <value>`	`--exclude-recent-search <value>`	`string` in the ms format	`3 days`	`undefined`

tip

How to ignore excludeRecentSearch for a single search?

What's the best way to add new trackers?

If you prefer to spread out repeat searches for a longer safety net, use excludeRecentSearch: "90 days" and excludeOlder: "450 days".

When running a search, this option excludes anything that has been searched more recently than this long ago. This option is only relevant in search mode or in daemon mode with searchCadence turned on.

Note

Search history is stored on a per-indexer basis.

`excludeRecentSearch` Examples (CLI)

cross-seed search -r 1day # only search for torrents that haven't been searched in the past day
cross-seed search --exclude-recent-search "2 weeks" # only search for torrents that haven't been searched in the past 2 weeks
cross-seed search --no-exclude-recent-search # disables/overrides the excludeRecentSearch value in config.js

`excludeRecentSearch` Examples (Config file)

excludeRecentSearch: "1 day",

excludeRecentSearch: "2 weeks",

`action`*

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`action`	`-A <save/inject>`	`--action <save/inject>`	`save`/`inject`	`inject`	`save`

cross-seed can either save the found cross-seeds, or inject them into your client. If you use inject, you will need to set up your client. Read more in the Injection tutorial.

`action` Examples (CLI)

cross-seed search -A inject
cross-seed search --action save

`action` Examples (Config file)

action: "save",

action: "inject",

`torrentClients`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`torrentClients`	N/A	`--torrent-clients <urls...>`	URL(s)	`[]`	`[]`

Multiple Clients

cross-seed will never inject a torrent into a client if it exists in another.

If there similar torrents to source from in multiple clients, the order of the clients defined will be used as the priority.

Injections from useClientTorrents will be injected into the client the match was sourced from or the first valid non-readonly client with the first valid linkDir. Custom routing is not supported at this time.

Injections from dataDirs will be injected into the first valid client with the first valid linkDir. Custom routing is not supported at this time.

The URL(s) of your torrent client that cross-seed will source torrents from prefixed by their client type. If desired, you can use readonly: after the prefix to use a client to source cross seeds from but not inject into.

tip

You may need to urlencode your username and password if they contain special characters

cross-seed must share the same user and group permissions as the torrent clients to prevent errors.

You may need to use cross-seed clear-client-cache if you change torrent clients or modify the torrents in client

`qbittorrent`

The url of your qBittorrent Web UI prefixed with qbittorrent: qbittorrent:http://user:pass@localhost:8080

`deluge`

The url of your Deluge JSON-RPC Interface prefixed with deluge: deluge:http://:pass@localhost:8112/json

`transmission`

The url of your Transmission RPC Interface prefixed with transmission: transmission:http://user:pass@localhost:9091/transmission/rpc

`rtorrent`

The url of your rTorrent XMLRPC interface prefixed with rtorrent: rtorrent:http://user:pass@localhost:8080/RPC2 or rtorrent:http://user:pass@localhost:8080/rutorrent/plugins/httprpc/action.php

info

If you use Sonarr or Radarr, cross-seed is configured the same way. ruTorrent installations come with this endpoint configured, but naked rTorrent does not provide this wrapper. If you don't use ruTorrent, you'll have to set up the endpoint yourself with a webserver.

tip

If you use HTTP Digest Auth on this endpoint (recommended), then you can provide credentials in the following format: http://username:password@localhost/rutorrent/RPC2

`torrentClients` Examples (CLI)

cross-seed search --torrent-clients rtorrent:http://rutorrent/rutorrent/RPC2
cross-seed search --torrent-clients rtorrent:http://user:pass@localhost:8080/RPC2 qbittorrent:http://user:pass@localhost:8080
cross-seed search --torrent-clients rtorrent:readonly:http://user:pass@localhost:8080/RPC2 qbittorrent:http://user:pass@localhost:8080

`torrentClients` Examples (Config file)

torrentClients: ["deluge:http://:pass@localhost:8112/json"],

torrentClients: [
	"qbittorrent:http://user:pass@localhost:8080",
	"deluge:http://:pass@localhost:8112/json",
	"transmission:readonly:http://user:pass@localhost:9091/transmission/rpc",
	"rtorrent:http://user:pass@localhost:8080/RPC2",
],

`notificationWebhookUrls`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`notificationWebhookUrls`	N/A	`--notification-webhook-urls <urls...>`	URL[]	`[]`	`[]`

cross-seed will send a POST request to these URLs with the following payload:

POST notificationWebhookUrl
Content-Type: application/json

{
  "title": "string",
  "body": "string",
  "extra": {
    "event": "RESULTS | TEST",
    "name": "string",
    "infoHashes": ["string"],
    "trackers": ["string"],
    "source": "announce | inject | rss | search | webhook",
    "result": "SAVED | INJECTED | FAILURE",
    "paused": false,
    "decisions": ["string"],
    "searchee": {
      "category": "string",
      "tags": ["string"],
      "trackers": ["string"],
      "length": 123,
      "clientHost": "string",
      "infoHash": "string",
      "path": "string",
      "source": "torrentClient | torrentFile | dataDir | virtual"
    }
  }
}

Currently, cross-seed only sends the "RESULTS" and "TEST" events. In the future it may send more. This payload supports both apprise and Notifiarr.

`notificationWebhookUrls` Examples (CLI)

cross-seed daemon --notification-webhook-urls http://apprise:8000/notify http://apprise:8001/notify

`notificationWebhookUrls` Examples (Config file)

notificationWebhookUrls: ["http://apprise:8000/notify", "http://apprise:8001/notify"],

`host`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`host`	N/A	`--host <host>`	`host/ip`	`0.0.0.0`	`0.0.0.0`

In Daemon Mode, cross-seed runs a webserver listening for a few types of HTTP requests. You can use this option to change the host to bind to and listen on.

Unless you have an address on your interface you wish NOT to listen on, or have a special set of networking requirements, you likely DO NOT need to change the host from "0.0.0.0".

`host` Examples (CLI)

cross-seed daemon --host 192.168.1.100

`host` Examples (Config file)

host: "1.3.3.7",

`port`*

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`port`	`-p <port>`	`--port <port>`	`number`	`2468`	`2468`

In Daemon Mode, cross-seed runs a webserver listening for a few types of HTTP requests. You can use this option to change the port it listens on.

`port` Examples (CLI)

cross-seed daemon --port 3000
cross-seed daemon -p 3000

`port` Examples (Config file)

port: 3000,

`rssCadence`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`rssCadence`	N/A	`--rss-cadence <cadence>`	`string` in the ms format	`30 minutes`	`undefined`

tip

How to cross seed new releases as soon as they are uploaded?

How to trigger ahead of schedule?

In Daemon Mode, with this option enabled, cross-seed will run periodic RSS searches on your configured indexers to check if any new uploads match torrents you already own. Setting this option to null, or not specifying it at all, will disable the feature.

tip

There is a minimum cadence of 10 minutes. We recommend keeping it at a relatively low number (10-30 mins) because if an indexer has a high frequency of new uploads, keeping the number low will make sure cross-seed gets a chance to see each new upload.

`rssCadence` Examples (CLI)

cross-seed daemon --rss-cadence 10min

`rssCadence` Examples (Config file)

rssCadence: null, // disable the RSS feature

rssCadence: "10 minutes",

rssCadence: "20min",

`searchCadence`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`searchCadence`		`--search-cadence <cadence>`	`string` in the ms format	`1 day`	`undefined`

tip

How to trigger a search on download completion?

How to trigger ahead of schedule?

warning

There is no reason to increase this value above the minimum of 1 day. This will negatively affect searchLimit, delay your searches, and cause them to be bunched up rather than spread out. If you wish to control the frequency in which torrents are searched, use the excludeOlder and excludeRecentSearch options instead.

In Daemon Mode, with this option enabled, cross-seed will run periodic searches of your torrents (respecting your includeEpisodes, includeNonVideos, excludeOlder, and excludeRecentSearch settings).

`searchCadence` Examples (CLI)

cross-seed daemon --search-cadence "2 weeks"
cross-seed daemon --search-cadence "2w"

`searchCadence` Examples (Config file)

searchCadence: null, // disable the periodic search feature

searchCadence: "2w",

searchCadence: "4 weeks",

`apiKey`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`apiKey`		`--api-key <key>`	`string`	`undefined`	`undefined`

info

apiKey is disabled in the config file by default, if you want to specify a key set it to a valid key (24 character min).

To find your generated API key, run the cross-seed api-key command. The API key can be included with your requests in either of two ways:

# provide api key as a query param
curl -XPOST localhost:2468/api/webhook?apikey=YOUR_API_KEY --data-urlencode ...
# provide api key as an HTTP header
curl -XPOST localhost:2468/api/webhook -H "X-Api-Key: YOUR_API_KEY" --data-urlencode ...

`apiKey` Examples (CLI)

cross-seed daemon --api-key <key> # will require auth on requests

`apiKey` Examples (Config file)

apiKey: undefined,
apiKey: "abcdefghijklmn0pqrstuvwxyz",

`snatchTimeout`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`snatchTimeout`		`--snatch-timeout <timeout>`	`string` in the ms format	`30 seconds`	`30 seconds`

warning

A single search (for which there could be thousands) could require multiple snatches which happens sequentially. Similarly, a single rss event could have thousands of candidates to parse through. Setting this value too high or to undefined could indenfintely delay how quickly cross-seed can process candidates for no real benefit.

This option applies to any snatch (download) of a .torrent file via Torznab. If a response is not given in the amount of time specified then it will consider the snatch as failed.

`snatchTimeout` Examples (CLI)

cross-seed daemon --snatch-timeout "15s"
cross-seed search --snatch-timeout "30s"

`snatchTimeout` Examples (Config file)

snatchTimeout: undefined, // disable the snatch timeout (http default)

snatchTimeout: "30s",

snatchTimeout: "15s",

`searchTimeout`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`searchTimeout`		`--search-timeout <timeout>`	`string` in the ms format	`2 minutes`	`2 minutes`

warning

cross-seed may need to perform thousands of searches in a single run. Setting this value too high or to undefined could indenfintely delay how quickly cross-seed can process candidates for no real benefit.

This option applies to any search via Torznab. If the search response is not given in the amount of time specified then it will consider the search failed.

`searchTimeout` Examples (CLI)

cross-seed daemon --search-timeout "20s"
cross-seed search --search-timeout "45s"

`searchTimeout` Examples (Config file)

searchTimeout: undefined, // disables searchTimeout (http default)

searchTimeout: "60s",

searchTimeout: "20s",

`searchLimit`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`searchLimit`		`--search-limit <number>`	`number`	`400`	`undefined`

Stop searching after the number of search queries meets the number specified.

Read more

`searchLimit` Examples (CLI)

cross-seed daemon --search-limit 50
cross-seed search --search-limit 150

`searchLimit` Examples (Config file)

searchLimit: undefined, // disable search count limits

searchLimit: 150,

`flatLinking`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`flatLinking`	N/A	`--flat-linking`	`boolean`	`false`	`false`

Be Advised

qBittorrent users using an external program or script, such as qbit_manage, will need extra considerations for flatLinking: false.

Read more about specific usage

With false, cross-seed will link any matches to a tracker-specific folder inside of linkDirs (if set). This prevents cross seeds from conflicting with each other. The tracker subfolder names will correspond to the ones set in Prowlarr/Jackett and autobrr, ensure they match to prevent multiple folders for a single tracker. Set this to true to use the flat-folder style linking previously used in v5.

With flatLinking: false (default):

linkDir/
├─ TrackerA/
│  ├─ Movie.mkv
├─ TrackerB/
│  ├─ Movie.mkv
│  ├─ Show S01/
│  |  ├─ Show S01E01.mkv

With flatLinking: true:

linkDir/
├─ Movie.mkv   <--- Both TrackerA and TrackerB cross seeds share the same file
├─ Show S01/
│  ├─ Show S01E01.mkv

`flatLinking` Examples (CLI)

cross-seed search --flat-linking

`flatLinking` Examples (Config file)

flatLinking: true,

flatLinking: false,

`blockList`

Config File Name	CLI Short Form	CLI Long Form	Format	Config Default	Fallback
`blockList`	N/A	`--block-list <strings>`	`string(s)`	`[]`	`[]`

tip

Use the blocklist on categories, tags, or trackers on torrents you do not want to cross seed from your torrent client.

cross-seed will exclude any of the files/releases from cross-seeding during the prefiltering done for each search/inject/rss/announce/webhook use. The full list of supported prefixes are:

name: A substring of the name inside the .torrent file or parsed name from path if data based.
nameRegex: Similar to name but uses your own custom regex.
folder: A substring of any parent folder in the path. Only applies to dataDirs searchees, not ones from your torrent client.
folderRegex: Similar to folder but uses your own custom regex on the entire parent path.
category: Deluge labels are considered categories. "category:" blocklists torrents without a category.
tag: Transmission and rTorrent labels are considered tags. "tag:" blocklists torrents without a tag.
tracker: If the announce url (from client/.torrent) is https://user:pass@tracker.example.com:8080/announce/key you must use host tracker.example.com:8080.
infoHash: Blocklist the torrent that matches this infohash (case-insensitive).
sizeBelow: Blocklist searchees with a size in bytes below this number.
sizeAbove: Blocklist searchees with a size in bytes above this number.

danger

The regex (ECMAScript flavor) options are for advanced users only. Do not use without rigorous testing as cross-seed is unable to perform any checks. Use at your own risk.

`blockList` Examples (Config file)

blockList: [
    "name:Release.Name",
    "name:-excludedGroup",
    "name:x265",
    "nameRegex:[Rr]elease[.\s][Nn]ame",
    "folder:folderName",
    "folderRegex:folder\d+",
    "category:icycool",
    "category:",
    "tag:everybody",
    "tag:",
    "tracker:tracker.example.com:8080",
    "infoHash:3317e6485454354751555555366a8308c1e92093",
    "sizeBelow:12345",
    "sizeAbove:98765",
],

Options on the command line​

Options in the config file​

All options​

delay*​

delay Examples (CLI)​

delay Examples (Config file)​

torznab*​

Finding your Torznab URLs​

torznab Examples (CLI)​

torznab Examples (Config file)​

sonarr​

Finding your Sonarr URL​

sonarr Examples (CLI)​

sonarr Examples (Config file)​

radarr​

Finding your Radarr URL​

radarr Examples (CLI)​

radarr Examples (Config file)​

useClientTorrents​

useClientTorrents Examples (CLI)​

useClientTorrents Examples (Config file)​

torrentDir​

torrentDir Examples (CLI)​

torrentDir Examples (Config file)​

outputDir​

outputDir Examples (CLI)​

outputDir Examples (Config file)​

dataDirs​

dataDirs Examples (CLI)​

dataDirs Examples (Config file)​

maxDataDepth​

maxDataDepth Examples (CLI)​

maxDataDepth Examples (Config file)​

linkCategory​

linkCategory Examples (CLI)​

linkCategory Examples (Config file)​

duplicateCategories​

duplicateCategories Examples (CLI)​

duplicateCategories Examples (Config file)​

linkDirs​

linkDirs Examples (CLI)​

linkDirs Examples (Config file)​

linkType​

linkType Examples (CLI)​

linkType Examples (Config file)​

matchMode​

matchMode Examples (CLI)​

matchMode Examples (Config file)​

skipRecheck​

skipRecheck Examples (CLI)​

skipRecheck Examples (Config file)​

includeSingleEpisodes​

includeSingleEpisodes Examples (CLI)​

includeSingleEpisodes Examples (Config file)​

seasonFromEpisodes​

seasonFromEpisodes Examples (CLI)​

seasonFromEpisodes Examples (Config file)​

autoResumeMaxDownload​

autoResumeMaxDownload Examples (CLI)​

autoResumeMaxDownload Examples (Config file)​

ignoreNonRelevantFilesToResume​

ignoreNonRelevantFilesToResume Examples (CLI)​

ignoreNonRelevantFilesToResume Examples (Config file)​

includeNonVideos​

includeNonVideos Examples (CLI)​

includeNonVideos Examples (Config file)​

fuzzySizeThreshold​

fuzzySizeThreshold Examples (CLI)​

fuzzySizeThreshold Examples (Config file)​

excludeOlder​

excludeOlder Examples (CLI)​

excludeOlder Examples (Config file)​

excludeRecentSearch​

excludeRecentSearch Examples (CLI)​

excludeRecentSearch Examples (Config file)​

action*​

action Examples (CLI)​

action Examples (Config file)​

torrentClients​

qbittorrent​

Options on the command line

Options in the config file

All options

`delay`*

`delay` Examples (CLI)

`delay` Examples (Config file)

`torznab`*

Finding your Torznab URLs

`torznab` Examples (CLI)

`torznab` Examples (Config file)

`sonarr`

Finding your Sonarr URL

`sonarr` Examples (CLI)

`sonarr` Examples (Config file)

`radarr`

Finding your Radarr URL

`radarr` Examples (CLI)

`radarr` Examples (Config file)

`useClientTorrents`

`useClientTorrents` Examples (CLI)

`useClientTorrents` Examples (Config file)

`torrentDir`

`torrentDir` Examples (CLI)

`torrentDir` Examples (Config file)

`outputDir`

`outputDir` Examples (CLI)

`outputDir` Examples (Config file)

`dataDirs`

`dataDirs` Examples (CLI)

`dataDirs` Examples (Config file)

`maxDataDepth`

`maxDataDepth` Examples (CLI)

`maxDataDepth` Examples (Config file)

`linkCategory`

`linkCategory` Examples (CLI)

`linkCategory` Examples (Config file)

`duplicateCategories`

`duplicateCategories` Examples (CLI)

`duplicateCategories` Examples (Config file)

`linkDirs`

`linkDirs` Examples (CLI)

`linkDirs` Examples (Config file)

`linkType`

`linkType` Examples (CLI)

`linkType` Examples (Config file)

`matchMode`

`matchMode` Examples (CLI)

`matchMode` Examples (Config file)

`skipRecheck`

`skipRecheck` Examples (CLI)

`skipRecheck` Examples (Config file)

`includeSingleEpisodes`

`includeSingleEpisodes` Examples (CLI)

`includeSingleEpisodes` Examples (Config file)

`seasonFromEpisodes`

`seasonFromEpisodes` Examples (CLI)

`seasonFromEpisodes` Examples (Config file)

`autoResumeMaxDownload`

`autoResumeMaxDownload` Examples (CLI)

`autoResumeMaxDownload` Examples (Config file)

`ignoreNonRelevantFilesToResume`

`ignoreNonRelevantFilesToResume` Examples (CLI)

`ignoreNonRelevantFilesToResume` Examples (Config file)

`includeNonVideos`

`includeNonVideos` Examples (CLI)

`includeNonVideos` Examples (Config file)

`fuzzySizeThreshold`

`fuzzySizeThreshold` Examples (CLI)

`fuzzySizeThreshold` Examples (Config file)

`excludeOlder`

`excludeOlder` Examples (CLI)

`excludeOlder` Examples (Config file)

`excludeRecentSearch`

`excludeRecentSearch` Examples (CLI)

`excludeRecentSearch` Examples (Config file)

`action`*

`action` Examples (CLI)

`action` Examples (Config file)

`torrentClients`

`qbittorrent`