An NZB is a small XML file that lists which Usenet articles belong to a release and how they fit together. It is not the movie, archive, or binary itself. Think of it as a shopping list of Message-IDs your downloader hands to your provider over NNTP. That indirection is why modern Usenet feels like a queue in SABnzbd or NZBGet instead of manual browsing in text newsgroups.
If you are coming from torrents, the habit to unlearn is treating the pointer as the payload. Indexers discover and publish NZBs; providers complete by returning the articles those files reference. When either side fails, you see missing articles, repair errors, or unpack failures that look similar in the history pane until you read the detail line.
NZB in one sentence
Each entry in an NZB maps to a specific article on a Usenet server, with offsets and filenames so the client can reassemble split RAR volumes in order. Metadata blocks may include subject tags, sizes, and grouping hints indexers use when they generate the file. The format is intentionally simple so automation can fetch without a human clicking through group headers.
Your provider still has to retain every listed Message-ID for the download to finish. A perfect NZB against an empty or takedown-hit spool produces the same failure as a broken NZB against a healthy spool. Completion is a provider plus indexer quality story, not an NZB magic trick.
NZB vs browsing newsgroups
Historically, users read newsgroups in a newsreader, selected attachments, and downloaded piece by piece. NZBs collapse that workflow into one import action. Easynews-style web search hides NZB complexity behind a browser UI, but the underlying fetch is still articles on a backbone.
Automation stacks prefer explicit NZBs. Sonarr and Radarr ask indexers for candidates, grab an NZB, and pass it to SABnzbd. That path scales because APIs replace eyeballing group noise. For role clarity across the three layers, read provider vs indexer vs newsreader.
What you still need besides NZBs
A complete beginner setup is provider plus downloader plus usually an indexer. The provider stores and serves articles. The downloader reads the NZB and orchestrates fetch, verify, repair, and unpack. The indexer maintains a searchable map from release names to fresh NZBs.
Free public indexers can work for light use; serious automation often pays for stability and API limits. None of that replaces SSL NNTP credentials from a retention-strong primary. If you wonder whether indexers are mandatory for your style, see do I need an NZB indexer.
Where NZBs come from
Indexers crawl or ingest headers, normalize release names, and emit NZBs when they trust a post set. Scene naming, obfuscation, and reposts mean the same show can appear as multiple NZBs with different ages and completeness hints. Prowlarr deduplicates across indexers so you are not manually picking files in a browser for every grab.
Binsearch and similar tools sit closer to raw header search; they can find posts indexers missed but require more judgment. For a deeper pipeline picture, follow how Usenet search works and pick stable sources from best NZB indexers.
Common failures that look like “bad NZBs”
Stale NZBs point at articles removed after the indexer snapshot. Wrong priorities can skip your fill server when you need it. Incomplete primary retention shows up as missing articles before PAR ever runs. Password-protected or mislabeled posts fail at unpack with a different error string than true missing segments.
Work through diagnosis before buying more providers. Read SABnzbd history for missing articles versus repair failure versus unpack failure. The playbook in avoid incomplete NZBs separates indexer ghosts from backbone gaps. A week of notes beats guessing.
FAQ
- Is an NZB a torrent?
- No. Torrents swarm peer to peer; NZBs list Usenet articles your provider serves over NNTP.
- Do I pay for NZBs separately from my provider?
- Often yes. Indexer memberships and provider subscriptions are different bills for different jobs.
- Are NZBs safe to download?
- Treat them like untrusted metadata. Import from indexers you trust; scan extracted content with the same hygiene you use for any download.
- Can my provider give me NZBs?
- Some bundles include search or curated lists, but automation still usually relies on dedicated indexers with APIs.