How does ComfyUI work?
Hi everyone!
I recently got an RTX 5070 and I've been hearing a lot about ComfyUI, but I still don't fully understand everything it's capable of.
I'm a graphic designer, and I'd love to use AI to improve my workflow. For example, I'd like to create realistic mockups, generate custom illustrations, typography, textures, and other design assets.
One thing I'm especially interested in is whether it's possible to train or guide models using a library of my own designs so the generated images follow the same visual style and aesthetic as my work. Is that something ComfyUI can do? If so, what's the best approach (LoRAs, Flux, ControlNet, IPAdapter, etc.)?
I'd also love to know:
- What are the most useful workflows you recommend?
- What are some things ComfyUI can do that other UIs can't?
- Is an RTX 5070 enough to take advantage of most of its features?
- If you were starting today, what would you learn first?
I'm trying to figure out if it's worth investing the time to learn ComfyUI and how I can get the most out of my GPU.
Thanks!