// the take
Spotifast is a native Rust/egui Spotify client for Linux, macOS, and Windows that plays music either through the open-source librespot library or through a separate "Soloist" player component, layering on Spotify Connect (including Cast and Sonos), local file playback, lyrics, Winamp skins, and a MilkDrop visualizer. It's for Spotify Premium subscribers who want something lighter and faster than the official Electron client and are fine with the legal/stability gray area that comes with any unofficial Spotify client.
Actual native UI via egui instead of a browser wrapper, so it really is lighter than the official Electron client, not just in marketing copy. Feature coverage goes well past basic playback: Spotify Connect with Cast and Sonos targets, local file playback alongside streamed tracks, and lyrics that fall back to LRCLIB when Spotify's own source is unavailable. The Winamp skin engine and MilkDrop port are non-trivial pieces of work, not just a gimmick slapped on top. Release cadence is real and sustained (release notes through v0.16.0) with packaging for Homebrew, AUR, Flatpak, Nix, and MSI, which says more about maintenance health than any CI badge would.
Playback is Premium-only and runs through an unofficial reimplementation of Spotify's protocol (librespot) or a separate Soloist component whose builds need periodic renewal and re-verification — that's a dependency on Spotify not changing something server-side, entirely outside the maintainer's control, and the project itself has a doc page addressing account-ban risk because this is a real concern, not a hypothetical. The Soloist audio routing is described as a set of platform-specific hacks (redirect output, apply volume after visualizers, restore on mute) that is inherently fragile and already has gaps: no audible EQ on Windows Soloist playback, no MilkDrop on Windows ARM. Setup also requires each user to register their own Spotify developer app and client ID, which is friction a consumer music app shouldn't normally impose. Nothing here is a stable 1.0 yet, so the protocol-breakage risk compounds with ordinary pre-1.0 churn.
We read the README, the directory tree and a key source file in November 2026. Nobody pays to be listed. How we judge →