Feeds Fun review: an open-source, AI-tagging RSS reader that scores and filters feed entries automatically. See features, self-hosting steps, and alternatives.
Feeds Fun is a free, open-source RSS and Atom reader built around one core idea: instead of showing every subscribed feed as one long chronological list, it uses AI-generated tags and user-written scoring rules to automatically surface the entries that matter most to a given reader and quietly demote or hide the rest.
It was created and is maintained by a single independent developer rather than a company, and it is distributed under the permissive BSD 3-Clause license, with both a public hosted instance and a self-hostable Docker deployment available to anyone who wants full control over their data and infrastructure.
The heart of Feeds Fun is its 'tag processor' architecture, which includes simple deterministic processors, such as tagging entries by their source domain or importing tags already present in the original feed, alongside an LLM-based processor that sends entry content to OpenAI or Google Gemini models to generate rich, topical tags automatically.
On top of these tags, users define scoring rules, for example adding points when a certain tag is present and subtracting points for another, which the application uses to continuously rank incoming stories, effectively building a personalized, transparent recommendation system rather than a black-box algorithmic feed.
Feeds Fun itself is entirely free, whether used via the hosted instance at feeds.fun or self-hosted through Docker Compose, and there is no paid subscription tier for the software.
The only cost a user might incur is indirect: because the AI tagging feature calls OpenAI's or Google Gemini's APIs using the user's own API key, any usage charges for those AI calls are billed directly by the AI provider to the user, separate from Feeds Fun itself.
Yes, Feeds Fun is fully open source and free, whether you use the hosted instance at feeds.fun or self-host it yourself; the only potential cost is usage fees from your own OpenAI or Gemini API key for AI tagging.
Feeds Fun is created and maintained by an independent developer known online as Tiendil, based in Germany; it is a solo open-source project rather than a venture-backed company.
You only need one if you want to use the AI-powered tagging feature; the rest of the reader, including feed subscriptions and manual tags, works without an AI key.
Yes, Feeds Fun can be self-hosted via Docker Compose with a single-user profile for personal use or a multi-user profile for small groups, and it requires a PostgreSQL database.
Feeds Fun is released under the BSD 3-Clause license, a permissive open-source license that allows free use, modification, and redistribution.
Unlike commercial cloud readers, Feeds Fun is free, open source, and self-hostable, and its core differentiator is a built-in LLM-based tagging and rule-scoring system for automatically ranking entries rather than relying on manual folders or keyword filters alone.