NZBHydra2: One Search Box for All Your Indexers

If you run more than two or three NZB indexers, you know the real cost isn’t the subscriptions — it’s the maintenance. Every indexer has to be configured separately in Sonarr, Radarr, and Lidarr; every API key rotation, every new indexer, every limit change happens three times. NZBHydra2 collapses all of that into a single newznab endpoint that fans your queries out to every configured indexer simultaneously and returns merged, deduplicated results.

What NZBHydra2 does

NZBHydra2 is a meta-search aggregator written in Java. You point your automation tools — Sonarr, Radarr, Lidarr, and the rest — at one URL, and it handles distribution behind the scenes. It speaks standard newznab and torznab APIs, so anything that can talk to an indexer can talk to NZBHydra2.

Beyond raw aggregation, it translates media IDs. When Sonarr searches by TVDB ID, NZBHydra2 passes that identifier down to every indexer that supports it, improving match quality without you having to configure anything per-indexer.

What you also get is observability. NZBHydra2 logs every search and every download, tracks per-indexer response times and hit rates, and surfaces that data through a built-in stats dashboard. After a few weeks of use you have a clear picture of which indexers actually pull their weight.

Integration points

• Download managers: SABnzbd and NZBGet for direct NZB delivery; torrent-to-folder for torznab results
• Media automation: Sonarr, Radarr, Lidarr, Readarr — add NZBHydra2 as a custom indexer with your local URL and API key
• Torrent trackers via Jackett: add Jackett’s torznab endpoints as sources, NZBHydra2 merges them with your Usenet indexers in one result set
• Mobile apps: nzb360 and similar tools that speak newznab

Multi-user mode with per-user API keys and access controls is available if you share your instance.

Recent changes worth knowing about (April – June 2026)

The project has been on a steady release cadence. A few things that matter in practice:

Indexer grouping (v8.7.0) lets you organize the indexer selection dropdown into named groups. If you have a dozen indexers split between general NZB, anime-specific, and Blu-ray sources, you can now label them as such rather than scrolling through a flat list.

Saved searches with size and age filters (v8.8.2 / v8.8.3) — saved searches now store and display min/max file size and age constraints. If you re-saved searches before v8.8.2, they need one more re-save to pick up the new fields.

Accessibility: dark theme for dyschromatopsia (v8.8.0) — a dedicated dark theme tuned for color vision deficiencies, separate from the standard dark mode.

CPU usage fix in image cache (v8.8.1) — a bug caused the image cache to spin the CPU. If you are on a version below v8.8.1 and notice elevated CPU, update.

Indexer list housekeeping (v8.8.2) — animetosho was removed from the built-in indexer list; ameNZB and BlurayNZB were added. The built-in list is a convenience; it does not stop you from adding any newznab-compatible indexer manually.

Season/episode data loss fix (v8.7.1) — a regression introduced in v8.7.0 dropped season and episode information during TV searches. Fixed the same day; if you were briefly on v8.7.0 and saw unexpected search results for TV content, that was the cause.

Running it

NZBHydra2 ships as a self-contained Java application and as a Docker image. Most people run it in Docker alongside their other stack components. There is no external database dependency — it uses an embedded H2 database by default.

The web UI runs on port 5076. First-run setup walks you through adding indexers and connecting a downloader. The newznab URL you give to Sonarr and the rest will be http://<host>:5076/api with the API key from the NZBHydra2 settings page.

Configuration lives in a single nzbhydra.yml file, which makes it straightforward to back up or migrate to a new host.

Where it sits in the stack

NZBHydra2 does not replace indexers — it multiplexes them. The more quality indexers you feed it, the better your hit rate. It also does not replace a download manager; it passes NZBs through to SABnzbd or NZBGet. Think of it as the routing layer between your media automation tools and the indexer landscape.

For anyone running multiple indexers without an aggregator, the drop in configuration overhead alone justifies the setup time — one place to manage API keys, limits, and indexer health instead of one per automation tool. The stats dashboard is a side benefit that tends to surface indexer problems — outages, throttling, degraded search quality — before they cause visible gaps in your library.

Project repository: github.com/theotherp/nzbhydra2