The name suggests what is about. It is randomizer prompt maker (characters/scenery/action/clothes/color/etc.) put into a image and then a video. After a HUGE work in progress I'm satisfied with the results. Yeah some results look wonky but it just the randomizer being random. Produces more NSFW than SFW. <s>I adjusted one of the nodes for</s><s> </s><s>wlsh</s><s>. </s> I updated to just plug in play (just download custom nodes and add the folder to the wildcards folder). I used a Nvidia 4090 RTX and produces a video in ~3 mins for a 5 sec generation after the first warmup generation that takes ~20 mins. Any tips/help would be cool. Have a nice day. Thx for reading and enjoy. :)

Randomizer Project (Txt2Img2Vid (Illustrious->Wan2)) (NSFW/SFW)

New Workflow. You don't need extra steps. Just plug in and play. Restarted from scratch. With help with ai for prompts. (Gemma 4 and Chatgpt/Ollama) Just download all custom nodes and put the newwild folder inside the "custom_nodes\comfyui-impact-pack\wildcards" folder. the new workflow is also inside the newwild folder. Enjoy.

Files with SHA256: 872a939d7a0974ae4afbef030f8a8100daf34c9e42cbaaa43920a67a58f72efe