← sources

EasyComfee V2: an open-source creative studio on top of ComfyUI

source: reddit (reddit:comfyui)published 2026-11-02 00:40ingested 2026-11-03 02:42

retained text

**EasyComfee V2: an open-source creative studio on top of ComfyUI**

For years ComfyUI felt like a fight: models that didn’t fit on my GPU, and more time wiring nodes than creating. So I built EasyComfee, a Higgsfield-style studio with Image, Video, VFX, Lip Sync, Music, Voice and Genjutsu, all running real ComfyUI workflows.

It reads your graphics card first. If the model fits, it runs locally for free. If it doesn’t, one click sends it to Comfy Cloud, with the price shown before you press Generate.

**New in V2:**

* A new interface with a side menu and a refined palette
* LTX at the core: keyframes on a timeline, 14 one-click VFX (3D render → video, day → night, remove people, swap character…), lip sync
* 3D render → Video Advanced mode, plus a Blender Bridge (MCP) to block a shot in Blender from any AI client
* Light video models for smaller GPUs: LTX-Video 2B, Wan 2.2, HunyuanVideo 1.5, Kandinsky 5
* Config Models grouped by task, with one-click installs and update checks
* Safe mode on by default, and Stop / Power buttons

Nothing is hidden: all 39 ComfyUI workflows are in the repo, and the Advanced tab opens ComfyUI itself. MIT licensed, contributions welcome.

[github.com/sintetizacom/easycomfee](http://github.com/sintetizacom/easycomfee)  
  
\#OpenSourceHiggsfield #ComfyUI #LTX

document fetched 2026-11-03 00:42 · canonical https://reddit.com/r/comfyui/comments/1xljadr/easycomfee_v2_an_opensource_creative_studio_on