Show HN: Demon – open-source real-time music diffusion engine, 25Hz local GPU
YO, I’m Ryan, lead author. I’ve been contributing open source generative audio stuff for a while now, audio reactive Comfy nodes, extended ACEstep support in Comfy, etc.. I just opened-sourced a new audio project that I've been working on for a few months and I want to tell y'all about it. WHAT IS IS DEMON: Diffusion Engine for Musical Orchestrated Noise This is StreamDiffusion but with audio instead of images, and ACEStep 1.5 instead of Stable Diffusion. It’s responsive enough that you can play it like an instrument, and remix in near real-time. I also distilled the ACEStep VAE: it’s faster at the expense of some quality. I also trained something like 200 lora/dora for ACEStep 1.5 and 1.5XL: I will release these in batches of 5 or 10 or something WHY IT IS Two reasons: 1) Making music is an inherently real-time activity 2) Why not bro SOME RUNTIME CAPABILITIES -Real-time
Why it failed
YO, I’m Ryan, lead author. I’ve been contributing open source generative audio stuff for a while now, audio reactive Comfy nodes, extended ACEstep support in Comfy, etc.. I just opened-sourced a new audio project that I've been working on for a few months and I want to tell
- Kind
- startup
- Status
- Shut down
- Inventor
- ryanontheinside
- Ended
- 2026
- Source
- hackernews