Runs the same prompt/settings through five models at once under the same seed. Orders the output of the images so they're reviewable on disk when sorted alphabetically (as opposed to split caused by the prefix+comfy suffix) that happens.

Useful for batch testing. Detailed notes within the document, so won't get overly wordy here.

Notes worth mentioning:

• It looks like a spaghetti mess, I did what I could to clean it while also keeping compatibility as high as possible for custom nodes and custom Use Anywhere linkage, this allows you to change/rewire anything to serve your needs without the annoyance of finding everything

• The naming schema for the output does cause a lot of this mess, but it is intentional

  • Uses the Run Counter node in Yellow to increment each time

  • Keeps the output as PREFIX_{RUN_#}_MODEL_#_#####

  • Allows you to sort them on disk alphabetically and scroll to review 1/2/3/4/5 without the output splitting based on the way Comfy normally handles the file saves

  • Unwire this if you don't have any need for it, I find it very useful to compare models, but it isn't critical infrastructure to generating here

• Test multiple models, or multiple LoRAs, or multiple LoRA weights, etc.

• SET LAST CLIP LAYER exists but is bypassed by default, I've seen newer Anima models recommend using it, so it's in place here in case you have need for it for a specific model