Plenio 1.0: free open-source AI song studio for ComfyUI ships with a player that renders videos

Vivid_Promise1700 · reddit · 2026-10-17

Plenio Music Production System 1.0 is out (Apache 2.0, fully free, no cloud): a local LLM (Gemma 4 natively, or any GGUF/LM Studio/Ollama) writes lyrics, YuE2 or MiniMax Music 3 performs them, and Plenio masters to -14 LUFS, exporting FLAC/MP3 with tags, cover art, lyric subtitles, PDF sheet music — and the whole project embedded in the FLAC for round-tripping into ComfyUI. Songs and workflows saved with 1.0 are guaranteed to work in every 1.x release.

New in 1.0, the Plenio Player adds:

Plenio Drop: MilkDrop-style GPU visuals, 16 presets, customizable via a short JSON file;
Live Play: an editable party playlist that picks up songs exported from ComfyUI instantly;
Song-to-video: visuals, karaoke lyrics, sheet music and title card rendered in Full HD/4K/square/vertical, several times faster than real time;
Share as a single HTML file — recipient double-clicks and gets lyrics, notation and visuals, no install or server;
11 fonts and 11 lyric effects (neon, glow, bounce, typewriter, beat-synced).

Also new: listen-timed lyric subtitles (SRT, half the words within 0.16 s), auto-filled tags, and one-switch seed controls. 26 native ComfyUI nodes, no extra Python packages, ComfyUI 0.37+, with 9 video tutorials and online demos.

Original post →

Plenio 1.0: free open-source AI song studio for ComfyUI ships with a player that renders videos

More from Multimodal