Anima Edit Image Editing Workflow

This is an image editing application based on Anima Base + Anima LLLite Inpainting + AnimaEditV1 + ZeroTwo Edit LoRA + Cosmos Reference Latent + InpaintModelConditioning. Users upload an original image and enter an editing instruction, such as "change to winter ski suit", "change background", "change to black dress", "add props", or "change to tea party scene". The system generates new edited results based on the original image structure and editing prompt.

The core advantage of this workflow is: Instead of redrawing a completely unrelated image, it performs targeted editing based on the original image. It standardizes the input image through ImageResize+, constructs side-by-side comparison editing inputs via AILab_ICLoRAConcat, enters partial repainting conditions through InpaintModelConditioning, and then uses Anima LLLite, AnimaEditV1, ZeroTwo Edit LoRA, and Cosmos Reference Latent to enhance editing stability. It is suitable for anime character costume changing, background replacement, prop addition, pose fine-tuning, inpaint, style unification, avatar retouching, AI photo editing, and secondary element material reprocessing.

When using, it is recommended to focus on controlling four items: The input image determines the original character and composition, the editing prompt determines what to change, the editing strength determines the magnitude of changes, and the random seed determines the reproducibility of the result.
There are multiple editing branches and multiple Preview outputs in the current workflow. When publishing the RunningHub application, it is recommended to converge into a single main output and hide template nodes like [75] / [77], showing users only "Upload image + Editing requirement + Editing strength + Random seed".

Before publishing, it is recommended to fix three points: First, the file name prefix of [101] SaveImage is still ComfyUI, it is recommended to change it to Anima_Edit_Result; second, the negative prompt is too short, it is recommended to add low quality, deformity, text, watermark, duplicate characters, wrong clothes, etc.; third, there are currently multiple AnimaEdit / ZeroTwo / Cosmos branches, and the frontend application is best to keep only one main result to avoid users not knowing which one is the final image.

🎁Claim RH Coins first, then go experience the workflow! Avatar in the upper right corner → Invitation code → Enter [rh-v1111] to instantly receive 1000 RH Coins, log in daily to get another 100 coins ~ For more ComfyUI workflows, tutorials, and gameplay, welcome to follow the official account (AIKSK), updated simultaneously on Douyin / Bilibili / Xiaohongshu / YouTube (AI-KSK) ✨