anima-base-v1.0-checkpoints
Model Information
Anima is a 2-billion-parameter text-to-image model developed in collaboration between CircleStone Labs and Comfy Org. It primarily focuses on anime concepts, characters, and styles, but can also generate a variety of other non-photorealistic content. The model is designed for producing illustrations and artistic images, and does not perform well in rendering realism.
It is trained on millions of anime images and about 800,000 non-anime art images. No synthetic data was used during the training process. The knowledge cutoff date for the anime training data is September 2025.
New feature: Try Turbo LoRA for greater stability and faster generation speeds.
Versions
Anima-Base
Pre-trained, unoptimized base model. Maximum flexibility, diversity, and style consistency.
Reference prompt: amber_(genshin_impact), 1girl, solo, long_hair,, bangs, , brown_hair, thighhighs, gloves, long_sleeves, , ribbon, cleavage, hair_between_eyes, brown_eyes, very_long_hair,, jacket, full_body, hair_ribbon, boots, shorts, grin, red_ribbon, short_shorts, floating_hair, thigh_boots, white_footwear, goggles, goggles_on_head, pouch, shrug_(clothing), vision_(genshin_impact), red_thighhighs, brown_shorts.Sit with your legs apart.
Steps: 40.
cfg: 4.5
Resolution: 1024*1536 or higher
Anima is a 2-billion-parameter text-to-image model developed in collaboration between CircleStone Labs and Comfy Org. It primarily focuses on anime concepts, characters, and styles, but can also generate a variety of other non-photorealistic content. The model is designed for producing illustrations and artistic images, and does not perform well in rendering realism.
It is trained on millions of anime images and about 800,000 non-anime art images. No synthetic data was used during the training process. The knowledge cutoff date for the anime training data is September 2025.
New feature: Try Turbo LoRA for greater stability and faster generation speeds.
Versions
Anima-Base
Pre-trained, unoptimized base model. Maximum flexibility, diversity, and style consistency.
Reference prompt: amber_(genshin_impact), 1girl, solo, long_hair,, bangs, , brown_hair, thighhighs, gloves, long_sleeves, , ribbon, cleavage, hair_between_eyes, brown_eyes, very_long_hair,, jacket, full_body, hair_ribbon, boots, shorts, grin, red_ribbon, short_shorts, floating_hair, thigh_boots, white_footwear, goggles, goggles_on_head, pouch, shrug_(clothing), vision_(genshin_impact), red_thighhighs, brown_shorts.Sit with your legs apart.
Steps: 40.
cfg: 4.5
Resolution: 1024*1536 or higher