Free ComfyUI templates chain local LLM lyrics into YuE2 song rendering, runs on 12GB VRAM
Vivid_Promise1700 · reddit · 2026-10-20
The same developer reposted their free ComfyUI music production templates (Plenio Music Production System, Apache 2.0) built around native YuE2 and MiniMax Music 3 nodes for local song generation.
Core workflow: a Song Brief feeds a local LLM writer (Gemma 4 or any GGUF / LM Studio / Ollama model); a Song Sheet node shows the editable style, lyrics and ABC score passed to YuE2; after rendering, EQ, loudness and tagged FLAC/MP3 export, with the workflow embedded in the FLAC for continued editing. New in 1.0.2: a creativity slider where 50 equals native defaults, adjusting YuE2 sampling (temperature, top-p, top-k).
Also includes covers, MiniMax Music 3, hand-written melodies and a mastering-only graph. Requires ComfyUI 0.37+, no extra Python packages; int8 YuE2 runs well on 12-16GB VRAM. YuE2 weights are CC BY-NC (non-commercial).
More from Multimodal
- 59 animations in a 43-min video: creator breaks down his AI agent workflow — MaxLenormand · 2026-10-20
- Dev generates an entire desert city with AI, says it would make great FPS maps — jasonkneen · 2026-10-20
- Feeding the whole Animator's Survival Kit to Opus 5.5 yields decent keyframe animation — MoonL88537 · 2026-10-20
- Glimpse launches as a visual thinking model claiming Astra-level reasoning benchmarks — LinusEkenstam · 2026-10-20
- ComfyUI lets you pipe images into an LLM — how do you do the same with audio? — PhilosopherSweaty826 · 2026-10-20
- AI-generated mecha battle video looks straight out of a movie — Key_Layer6250 · 2026-10-20