Wildcards v1.0 — Anima + SDXL + Krea 2 (One Workflow)

Fire and forget. My daily driver for Image Generation. Pick a model family, load your LoRA(s), raise the batch amount, hit Run.
Questions about the graph? Come ask on my Discord: https://discord.com/invite/67C7P5xtFa

The wildcards build a fresh prompt around your LoRA every time.

WHAT IT DOES:

One workflow, three families (Anima, SDXL/Pony/Illustrious, Krea 2). It takes your LoRA's trigger words and automatically mixes them with quality tags, poses, locations, lighting, clothing — whatever you set. Wildcards live right in the graph, so no extra .txt files needed.

Want full control? Flip "Full Manual" to true at the bottom of that family's box.

QUICK START:

• Install missing nodes with ComfyUI Manager (including Animora Family Switch), then restart ComfyUI.

• Download the models for the family you want — links and folders are in the info note on the canvas.

• Pick Anima, SDXL or Krea with the Family Switch on the bottom dashboard, and load your models.

• Add LoRAs with LoRA Manager. Tip: click the blue "L" at the top to open the visual LoRA browser and send one over in a click.

• Set seed, size and steps on the Dashboard, then Run. Images save to the Output group.

MAKE YOUR OWN WILDCARDS

Type options in curly braces: {red|blue|black} shirt — each run picks one.

REQUIRED CUSTOM NODES (+available in manager):

• Animora Family Switch: https://github.com/StudioAnimora/Animora-Family-Switch

• LoRA Manager: https://github.com/willmiao/ComfyUI-Lora-Manager

• Pixaroma: https://github.com/pixaroma/ComfyUI-Pixaroma

• Impact Pack: https://github.com/ltdrdata/ComfyUI-Impact-Pack

• rgthree: https://github.com/rgthree/rgthree-comfy

MODELS:

SFW Options listed. If wanting NSFW hit up .red

Anima

• anima-base-v1.0: https://huggingface.co/circlestone-labs/Anima/resolve/main/split_files/diffusion_models/anima-base-v1.0.safetensors

• qwen_3_06b_base: https://huggingface.co/circlestone-labs/Anima/resolve/main/split_files/text_encoders/qwen_3_06b_base.safetensors

• qwen_image_vae: https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors

SDXL

• Any SDXL, Pony or Illustrious checkpoint (interchangeable)

• SDXL: https://civarchive.com/models/133005/juggernaut-xl?modelVersionId=1759168

• Pony:https://civarchive.com/models/257749/pony-diffusion-v6-xl?modelVersionId=290640

• Illu: https://civarchive.com/models/795765/illustrious-xl

Krea 2

• Krea 2 Turbo: https://civarchive.com/models/2726029/krea-2-turbo-official-comfy-org-checkpoints-krea2

• qwen3vl_4b_fp8: https://comfy.org/p/supported-models/qwen3vl-4b-fp8-scaled/

• qwen_image_vae: https://huggingface.co/Comfy-Org/Qwen-Image_ComfyUI/resolve/main/split_files/vae/qwen_image_vae.safetensors

FAQ

Q: How do I get more wildcards?

A: Add your own in the graph with {option|option|option}. Stack as many as you like.

Q: Why does SDXL only need one download?

A: SDXL checkpoints already include the text encoder and VAE, so one file covers it.

Q: My LoRA is loaded but not working?

A: Make sure the LoRA matches the family you switched to (an SDXL LoRA won't work on Anima), and that its trigger words are in the trigger box.

KNOWN ISSUES

Random wildcards can sometimes clash. When writing your own, keep options broad rather than overly specific.

Custom nodes run on your own setup, so give them a quick look before running. Questions or tweak requests? Drop them in the comments!

CONTENT RATING

The demo wildcards are SFW. Each family's prompt follows its model's own rating tags, so if your model supports it, you can change the rating tag and edit the wildcard lists to suit your own content. You're responsible for following Civitai's content rules and your model/LoRA licenses.

Also on Stable Yogi: https://stableyogi.com/portal/resource.html?slug=wildcards-v1-0-anima-sdxl-krea2-workflow

More models and workflows: https://stableyogi.com/portal/models.html

Welcome to the latent space of Wildcards 🍀

Made by Animora (thank you Pixaroma for a wonderful pack.)