Feeds Fun

Feeds Fun is a news reader for people whose RSS subscriptions produce more articles than they want to read. It assigns tags automatically, then uses rules you define to score articles by topic. You can self-host it with Docker or use the hosted service at feeds.fun.

The scoring rules give you direct control over what appears first. You can favor subjects you follow, lower the score of topics you rarely read, or combine tags in a rule. Filters let you exclude articles with certain tags or show only those that match your interests. The reader also sorts by date and tracks articles you've already read.

AI tagging works with OpenAI and Google Gemini, which require an API key for their cloud services. It also supports other models through OpenAI-compatible or Gemini-compatible APIs, including custom endpoints. Self-hosting runs the reader on your own server; where AI processing happens depends on the model endpoint you choose. Fetching feeds requires network access.

Tags can come from simple rules, such as identifying a site's domain, as well as an LLM's analysis of an article. The reader normalizes generated tags to reduce inconsistencies that would otherwise make filters and scoring rules harder to reuse. Self-hosted deployments support a single reader or multiple users.

The hosted site includes curated feed collections for entrepreneurship and startups, computer science, mathematics and physics.