Anima Batch Output (5 Models, 1 Prompt) for ComfyUI — Anima (Diff Model) v1

Author notes from animaBatchOutput5Models1_animaDiffModelV1.json · selected release

KSamplers, left exposed so if you want set custom sampler or scheduler values on a specific model basis you can You can also unwire Steps, CFG, Seed Generator, etc. to set specific values here

Author notes from animaBatchOutput5Models1_animaDiffModelV1.json · selected release

Set each LoRA you want to run, what weight you want to run it at, or if you want to run LoRAs at all Useful for testing one model across multiple LoRAs, or testing different LoRA/weight combinations (which is why Power Lora Loader is used here)

Author notes from animaBatchOutput5Models1_animaDiffModelV1.json · selected release

The reason this mess exists above instead of in a subgraph is you may have need to change things, and I find digging stuff out of subgraphs annoying. Most of these don't require settings changes The noteable exception is the Something To String nodes are used to properly set/group the output into reviewable format: - Prefix -- `SPFlow/Anima_Batch/Run_` - Middle -- `Value from the Counter` - Suffix -- `Model_##` This makes each run follow the counter for ordering on your drive so that it doesn't eventually split 1/2/3/4/5 out as the output image counter increments, keeping them together so you have 5 images in a row, ordered correctly alphabetically, regardless of what order ComfyUI runs the flow in

Author notes from animaBatchOutput5Models1_animaDiffModelV1.json · selected release

CLIP and VAE included above so that if you want to flip this to another, non-Anima, Diff Model you can and load the appropriate CLIP/VAE (Z-Image, Flux, etc.) Usage: 1. Set your settings in the first box (Steps, CFG, Seed, Latent Image Size/Aspect Ratio, and Positive/Negative Prompt) 2. Set your models (5 different, or all the same) 3. Set your LoRAs (different, none, combination, useful for setting the same model five times and trying different LoRAs or different LoRA weights, or testing LoRA training bakes) 4. Send it, get your images, have them ordered in the order of the RUN COUNTER on disk alphabetically, instead of in the ever incrementing ComfyUI counter splitting each batch run

Author notes from animaBatchOutput5Models1_animaDiffModelV1.json · selected release

Output Images, set in name to group together when sorted alphabetically, using the Run Counter in the top portion. Example - Run 1: Run_1_Model_01_00001_ Run_1_Model_02_00001_ Run_1_Model_03_00001_ Run_1_Model_04_00001_ Run_1_Model_05_00001_ Run 2: Run_2_Model_01_00002_ Run_2_Model_02_00002_ Run_2_Model_03_00002_ Run_2_Model_04_00002_ Run_2_Model_05_00002_ Keeping your Model or LoRA reviews in order in your image preview program of choice

Author notes from animaBatchOutput5Models1_animaDiffModelV1.json · selected release

Set each Diffusion Model you want to run, or set the same and apply different LoRAs/LoRA weights below to test impact

Author notes from animaBatchOutput5Models1_animaDiffModelV1.json · selected release

Specifically included here because I've had no need for it with Anima, but I've seen more recent models request it (or if you want to swap to another Diff Model platform, it's available. They are set to Bypass by Default, but to change this either Right Click -> Bypass or select the node(s) you need and hit Ctrl+B If nothing else this space could serve as a nice placeholder for you to wire in whatever you feel like.