💀💀💀 READ THIS FIRST 💀💀💀

If you get a red error message, just download the missing nodes via Node Manager

Or just copy paste the error summary into Gemini, trust me they're your friends

Before going wild, you need to download these extra support files

so make sure to check the description of whichever ANIMA model you're using

For now, most models still need these, they aren't baked-in like ILLUS or PONY

• TEXT ENCODERS: wen_3_06b_base.safetensors place in ComfyUI/models/text_encoders

• VAE: qwen_image_vae.safetensors place in ComfyUI/models/vae

🌊🌊🌊

ANIMA handles eyes and hands pretty well on its own, you can mostly bypass the detailer

The default parameters are already solid, but you can change them if needed

The detailer is off by default, so just un-bypass the group to turn it on

There are only six nodes total

🌊🌊🌊

The Load LoRa chain fits up to 9 LoRAs

Don't forget to set the model strength and bypass the unused slots

🌊🌊🌊

Set your resolution in Empty Latent Image

🌊🌊🌊

Most ANIMA models play nice with my KSampler settings

Test it first and tweak as you go

🌊🌊🌊

This workflow works for ILLUS and PONY too, just bypass Load Clip and Load Vae nodes

Connect the cable straight from Load Checkpoint, It's literally drag-and-drop, don't overthink it