Anima is a 2 billion parameter text-to-image model created via a collaboration between CircleStone Labs and Comfy Org. It is focused mainly on anime concepts, characters, and styles, but is also capable of generating a wide variety of other non-photorealistic content. The model is designed for making illustrations and artistic images, and will not work well at realism.

It is trained on several million anime images and about 800k non-anime artistic images. No synthetic data was used for training. The knowledge cut-off date for the anime training data is September 2025.

This preview version is an intermediate model checkpoint. The model is still training and the final version will improve, especially for fine details and overall aesthetics.

Preview2 The preview2 version is a small upgrade to the first preview.

A significant part of the training is redone with different hyperparameters and techniques, designed to help make the model more robust to finetuning. It is trained for much longer at medium resolutions in order to acquire more character knowledge. A regularization dataset is introduced to improve natural language comprehension and help preserve non-anime knowledge. It has the same resolution limitations as the first preview. It is trained only briefly at 1024 resolution. Going much beyond this will cause the model to break down. This is a base model with no aesthetic tuning. It is designed to be wild and creative, with the maximum possible breadth of knowledge. It is not optimized to produce aesthetic or consistent images.