The best face swap I have used! Not PuLID! No LoRA Training Required. ComfyUI ZenID

[@ramn_]

Always the best content in your channel, thank you. Greetings from México.

[@Dunc4n1d4h0]

I feel like I used time travel machine and it is 2023 and we use IPAdapter with SDXL

[@ivanbelyakov]

There is no Workflow for SDXL in the 2025 folder, according to your link. I really wanted to download it (
In any case, thanks for your videos. Subscribed.

[@CodeCraftersCorner]

Sorry for that. I updated the folder. Can you check again and let me know if there are any issues.

[@zeltfor]

isn't this practically doing the same thing as ReActor but just more complicated? I first thought this was something brand new which allows us to generate a specific face during the generation process so it would allow us to generate something on the face like a nose ring or sunglasses for example, without them then being removed by the face swapping process. Now this just seems to do the exact same thing as ReActor, or am I mistaken?

[@PrivateMilo-d2s]

can it does multiple people?

[@abaj006]

Does this work with Flux Dev models?

[@teddygabarr]

with 6gb of vram this workflow does work?

[@rangorts]

is it better than pulid for realism?

[@MEGTON]

Please make a video on lora training locally

[@edmartincombemorel]

you don't ned grounding dino to segment the face, most of the time crop face with dlib model will work, it is way better for lowram gpus

[@francaleu7777]

🎉

[@tsoukkk]

No folder zen ID in folder 2025

[@CodeCraftersCorner]

Sorry for that. I updated the folder. Can you check again and let me know if there are any issues.

[@procrastonationforever5521]

Dude... This is creepy AF, lol! xD)))) Sharvin waifu? What a narcissist, huh?

[@UnclePapi_2024]

👏🐵👏

[@dirkrabe1201]

Sorry, I can't find a workflow in: 2025/ZenID folder

[@CodeCraftersCorner]

Sorry for that. I updated the folder. Can you check again and let me know if there are any issues.