Skip to content

Search and indexers

When you play a title, NZB-DAV searches every enabled provider, merges the results, removes duplicates, and hands the combined list to the filtering and ranking stage.

Provider types

You can enable any combination of three provider types. If none is enabled, NZB-DAV tells you so instead of searching.

NZBHydra2

NZB-DAV queries NZBHydra2's Newznab XML API. For episodes it prefers a TVDB id (then an IMDb id) so results are accurate; for movies it uses the IMDb id. If an id-based query returns nothing, NZB-DAV automatically retries by cleaned title so a missing or mismatched id doesn't leave you with zero results.

Prowlarr

NZB-DAV queries Prowlarr's native search API, which returns JSON. Because Prowlarr's native search binds ids inside the query text rather than as separate parameters, NZB-DAV embeds them as tokens — {tvdbid:…}, {imdbid:…}, {season:…}, {episode:…} — alongside the cleaned title.

Prowlarr contributes Usenet results only

NZB-DAV keeps only releases whose protocol is usenet. Torrent results from Prowlarr are silently dropped. A torrent-only indexer in your Prowlarr indexer list contributes nothing.

Set Prowlarr Indexer IDs to a comma-separated list to query specific indexers. Prowlarr returns no results if this list is empty, so specify at least one indexer ID.

Direct Newznab indexers

If you don't run Hydra or Prowlarr, connect directly to individual Newznab indexers. NZB-DAV queries them in parallel and includes each indexer's caps (advertised search capabilities) when building queries.

Two ways to configure them, both on the Indexers tab:

  • Popular indexers — built-in rows for NZB.su/NZB.life, NZBGeek, NZBFinder, NZBPlanet, DrunkenSlug, and DOGnzb. Enable one, enter its API key, done.
  • Manage Indexers — a dialog for adding any Newznab indexer from a larger preset catalog (23 well-known indexers) or a fully custom URL, and for testing, editing, enabling/disabling, and deleting them.

The Manage Indexers dialog validates each URL, re-fetches caps when you change a connection, and warns you before you disable your last enabled indexer.

How results are combined

flowchart LR
    H[NZBHydra2] --> M[Merge]
    P[Prowlarr] --> M
    D[Direct indexers] --> M
    M --> DD[De-duplicate by download link]
    DD --> R[Combined result list]
  • Each provider normalizes its results into a common shape: title, download link, size, indexer name, post date, and age.
  • De-duplication is by download link. The first occurrence of a link wins. A release that two providers return with different download URLs appears twice — this is intentional, because those are genuinely different downloads.
  • A result with no download link is dropped, because it can't be played.

Search caching

Searches are cached so that re-opening the same title is instant. The cache duration is the Cache duration setting (default 300 seconds; set to 0 to disable). The cache stores the raw, pre-filter results, so changing your filter or sort settings takes effect immediately without a new search. Clear it any time from the add-on's main menu (Clear Cache).

For the internal mechanics — the query planner, caps handling, and the exact result fields — see How it works → Search pipeline.