Skip to content

Indexers

Indexers are the search services streamarr.media queries to find releases — downloadable copies of movies, episodes, music, games, and books. Every automation feature that acquires media (Smart Play, favorites auto-download, quality upgrades, RSS sync) searches through the indexers configured here. The releases they return are downloaded by a download client.

Supported types

Type Protocol Typical use
Newznab Usenet indexer API NZB indexers
Torznab Newznab-compatible API for torrents Jackett, Prowlarr, native Torznab trackers

Adding an indexer

Navigate to Admin → Indexers and click Add Indexer. A two-step wizard opens.

Step 1: Connection

Field Description
Name A descriptive label
Type Newznab (Usenet) or Torznab (Torrent)
Host / URL e.g. https://indexer.example.com
API Key Your personal API key from the indexer
Use SSL / Verify SSL certificate Connection security options
Enabled Whether the indexer participates in searches
RSS sync Opt this indexer into the periodic RSS poll (see below)
Priority Lower numbers are queried first (default 25)

Click Test Connection. The test contacts the indexer, validates the URL and API key, and fetches its capabilities; on success you see "Connection successful — N categories found" and the wizard advances.

Step 2: Categories

Select the categories you want to search from the tree the indexer reported. For each selected category you assign:

  • Media Type — Movies, Series, Music, Games, Books, Audiobooks, or Other. Pre-filled from the standard Newznab ID ranges (2000s → Movies, 3000s → Music, 5000s → Series, 6000s → Games, 7000s → Books).
  • Languages and Resolutions (optional) — hints describing what this category contains.

Category hints improve scoring

Language and resolution hints are used as fallbacks: when a release title itself doesn't reveal its language or resolution, the category's hints are applied before scoring. This helps a lot with single-language indexer categories (e.g. a "Movies/German" category).

Editing an indexer

Edit on the indexer list reopens the wizard. Notes:

  • The API key is never sent back to the browser — leave the field empty to keep the existing key.
  • Continue without test skips the connection test; Reload categories re-fetches the capability tree while keeping your saved category assignments.
  • The last RSS sync time is shown on the connection step.

How a search works

Searches fan out concurrently to every enabled indexer whose categories match the media type, using a tiered, ID-first strategy (Sonarr/Radarr-style):

  1. Tier 0 — IMDb-ID search, no query text
  2. Tier 1 — title search, only if tier 0 returned nothing
  1. Tier 0 — TVDB/IMDb-ID search with season/episode parameters
  2. Tier 1 — title search fallback
  1. Tier 0 — structured search (artist + album / author + title)
  2. Tier 1 — combined text query fallback

Results are de-duplicated across tiers and indexers. Transient network errors are retried; a rate-limited indexer (HTTP 429) fails that search rather than hammering the service.

Release matching and scoring

Returned releases go through a strict pipeline before anything is grabbed:

  1. Rejection filters — video releases under 50 MB (samples), raw disc images (ISO/BDMV/VIDEO_TS), multi-season packs when a single episode is wanted, and releases published before the media's release date are dropped.
  2. Title matching — releases are matched against the media item by external ID first, then exact, normalized, title+year, and finally fuzzy title similarity.
  3. Quality/language scoring — each surviving release is scored using the library's Download Scoring Rules (per-library, under Libraries → library settings): weights for resolution, source, video and audio codec, trusted release groups (bonus) and blocked release groups (always scored 0), plus a language match bonus. Releases whose language is outside the library's allowed languages are rejected. Fresh releases get a small age bonus, very old ones a penalty.
  4. The highest-scoring release is dispatched to a matching download client and imported after completion.

RSS sync

RSS sync periodically polls the latest-releases feed of every indexer with the RSS sync toggle enabled, so new releases of monitored favorites are grabbed minutes after they appear instead of waiting for the next scheduled search.

  • Enable it globally in Admin → Settings under Favorites Automation ("RSS sync (poll indexers for new releases)"), and set the poll interval (default 15 minutes).
  • Enable it per indexer in the indexer's connection step. Indexers are polled in priority order.
  • Feed items are matched against monitored favorites by title. A match only triggers the normal search-match-score pipeline above — raw feed items are never downloaded directly, so a noisy feed cannot cause a false grab.

Note

RSS sync only results in downloads when favorites auto-download is also enabled in Favorites Automation. See System Settings for the surrounding automation switches.