Is it possible to integrate comfyui into Blender (no offense intended)
I know everyone might harbor a lot of malice towards AI, but there's no denying that we have entered such a chaotic era. I'm wondering if we can utilize the internal UI of Blender to use ComfyUI, given that both are open-source. Perhaps in the future, confui will only provide Blender with a plethora of functions conducive to artistic creation, such as audio generation, texture generation, and model generation, thereby liberating creators from the complex industrial pipeline. We can even utilize AI to batch process some repetitive operations. Think about it, it's like another form of renderer. Perhaps external interfaces can be provided to realize some forms of AI-assisted existing workflows, which can also bring more users to Blender. The UI of Blender is very beautiful, while ComfyUI is very outdated. Perhaps it's a good idea In this way, Blender may be able to integrate generative digital art, and the workflow of ComfyUI can be saved in the Blender asset browser, and in the future, large models can be utilized to process plugin code and so on
7 Comments
In fact, I did this a long time ago. The results are as follows:
github.com/AIGODLIKE/ComfyUI-BlenderAI-node
Guess why?
If you want to use comfy, it’s already out there. Go use it.
If you need all of those features to be pulled into blender, because it’s not easy enough for you to fake artistic skills, then that just hits a new level of absurdity.
I’m not advocating art forgery. I find it odd that you immediately equate my idea with forgery and use it to accuse me. I’m only asking whether it’s possible: AI’s ability to recognize diverse inputs supports generative art—not the direct theft you mention, but something more like a sensor. You can call it a bad idea, that’s fine. I’m just exploring possibilities. Even angry, I can’t ignore the elephant in the room; we should learn to coexist with it, not fight it. Anger, blame, or slander won’t change the fact that AI is already here. If the entire industry workflow is reshaped by AI, but tools refuse to adapt out of emotion and lose users, would you really want that? We should instead integrate AI into traditional workflows, creating new methods that free creators from repetitive work. Plugins and presets already speed up our work—do you call that stealing? Must every artist rebuild a computer from scratch just to create? Calm down. This era is far from perfect, but it’s here. Focus on how to live with the elephant, not how to defeat it.
No matter how angry you get, the reality is that AI has already been widely adopted in business. This is a change brought by AI itself, not by me. Of course I deserve criticism for raising this question, but think calmly: if creators eventually get used to working with AI, will our anger today still matter? There’s no denying that 3D is full of repetitive work—retopology, rigging, to name just two. I mentioned ComfyUI, but I really mean AI in any form. ComfyUI just seems to fit Blender’s node-based workflow better. If we can find a way of human-AI collaboration, the future might not be as bad as we think. Let’s be honest: CG art is essentially simulated art. Maps like diffuse, roughness, and height are all imitations of real-world photography. No one is inherently nobler than anyone else. Do you really expect praise just because you spent a month manually rigging a model? Tools are just tools. If AI only copies, it will never be taken seriously. So just relax.
if u don't want to use it, don't use it. Why do u have be skeptical about it. Asking everyone who proposes the idea to shut up. We have a platform like these, so people can give feature suggestions!!!
i have proposed a similar idea like these, but not comfy ui, an assistant integrated into blender, no 1 supported it. I know these is a good idea, but the 3d community is threatened by idea's like these. I will upvote these, but i don't think the developer would choose these, because no community support. So effort of developers won't put in the effort to develop it!!!!
Yes, I feel that people have lost their rationality in opposing AI, as if opposing AI is politically correct. Even though most of their current work is based on the achievements of predecessors, this is what I find very strange. When a new thing has already emerged, and there is no other option but to oppose it, is this state good? It's like a hysterical person who is unwilling to try to understand or acknowledge that this is already part of reality. AI actually has sufficient development value as a workflow in the early stages or as a work assistant. However, in the CG field, no one is willing to truly explore its integration with practical work. Instead, they act like religious believers who reject heretics. I think it may not be possible to really have constructive exchanges here