gitmyhub

v2

Go ★ 9.4k updated 1d ago

Minimalist and opinionated feed reader

Miniflux is a minimal self-hosted RSS reader that strips tracking from articles, connects to 25-plus services like Telegram and Notion, and runs as a single binary with only a PostgreSQL database required.

GoPostgreSQLDockersetup: moderatecomplexity 3/5

Miniflux is a self-hosted RSS and feed reader you run on your own server. It collects articles from websites that publish feeds in the Atom, RSS, or JSON Feed formats and puts them in one place, so you can read everything without visiting each site individually. You subscribe to blogs, news outlets, podcasts, or YouTube channels, and Miniflux checks for new content in the background on a regular schedule.

Privacy is a central concern. The software strips tracking codes from article URLs, removes invisible tracking pixels embedded in content, and routes YouTube videos through a privacy-friendly player. External scripts from third-party sites are blocked by default, so articles you read cannot report your activity back to advertisers or data collectors.

Miniflux connects to more than 25 external services, including Telegram, Slack, Notion, Instapaper, Wallabag, and Readwise Reader. You can send new articles to a read-later service, fire off a notification when a post goes live, or save bookmarks to a link manager, all through the settings panel. It also exposes a REST API and speaks the Fever and Google Reader protocols, so some existing mobile feed reader apps can connect to it directly.

Installation requires a server and a PostgreSQL database. Miniflux ships as a single compiled file with no additional dependencies, and official packages exist for Debian and RPM-based Linux systems. A Docker image is available for standard and ARM hardware. HTTPS certificates can be set up automatically through Let's Encrypt.

The interface is intentionally plain: no recommendation engine, no social layer, no visual clutter. It supports keyboard shortcuts for fast navigation, several light and dark themes, and full-text search across everything you have collected. The software is available in 20 languages. It uses very little memory and CPU even when tracking hundreds of feeds.

Where it fits