Anima Role Ref KV · v1.0

Generate new poses, outfits, expressions, and scenes from a character reference image and a text prompt, while keeping the character recognizable.

Required downloads: the model weights AND anima-role-ref-kv-comfyui-setup-v1.0.zip from this version’s attachments. The ZIP contains the required custom node, workflows, and installation guide. The weights alone are not sufficient; a standard LoRA Loader cannot run this reference workflow.

NSFW content was filtered out of the training data and was not used to train this release.

For ComfyUI · Requires Anima Base v1.0. Place the weights in ComfyUI/models/loras/. Download the required custom node and workflow from [Hugging Face](https://huggingface.co/AnticPan/Anima-Role-Ref-KV), restart ComfyUI, import the workflow, upload a reference, and Queue. The Anima Role Ref KV node loads the adapter itself.

Suggested settings: 28 steps, CFG 4, BF16, around 1 megapixel.

Try 1024 × 1024, 832 × 1216, or 1216 × 832; dimensions must be divisible by 16.

No trigger word is required.

Examples: left = reference image; right = generated result. Use a clear single-character reference and describe the desired changes. Difficult references or strong changes can reduce resemblance.

This is an independent modified derivative of Anima, not an official or endorsed CircleStone Labs product. Use is subject to the [CircleStone model license](https://huggingface.co/circlestone-labs/Anima-Base-v1.0-Diffusers/blob/main/LICENSE.md); rights are granted directly by CircleStone Labs.

The CircleStone Model is licensed by CircleStone Labs LLC under the CircleStone Non-Commercial License. Copyright CircleStone Labs LLC.

IN NO EVENT SHALL CIRCLESTONE LABS LLC BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.