Text-to-Image
Cosmos
draw-things
s4nnc
anima

dt-anima

Anima Turbo 1.0 weights in the Draw Things (s4nnc .ckpt) format, mirrored unmodified from the Draw Things model CDN for use by JoyFusion's on-device inference. The conversion and quantization were done by the Draw Things project; the file here is byte-identical to the upstream one.

File Variant Size SHA-256
anima_turbo_1.0_q8p.ckpt 8-bit 2,375,991,296 7e711bc61a4b6fd06cda32ccdc2fd702f5b30be91638dc9e38096836fda5864f

The shared components live in JoyFusionAI/dt-components:

File Role Size SHA-256
qwen_3_0.6b_q8p.ckpt Text encoder (Qwen3-0.6B, 8-bit) 767,352,832 fee6eca8ecbeb8f75fccf553968f6964d5191e3441d925cd89984503ab8a8d64
qwen_image_vae_f16.ckpt Qwen-Image VAE 254,500,864 701e7c46ed6c2fa8036543780317e06d264374f9b4dbfc22f27c0b3181bb988a

Anima is a 2B text-to-image model built on NVIDIA Cosmos, focused on anime, illustration and other non-photorealistic art. Upstream recommends around 1MP output, 4–8 sampling steps and CFG 1 for the Turbo checkpoint.

License

Anima is licensed by CircleStone Labs LLC under the CircleStone Labs Non-Commercial License v1.2 (see LICENSE.md), which permits use and distribution for non-commercial purposes only. See NOTICE for the required attribution notice. Any rights to use the model are granted directly by CircleStone Labs LLC to you under that license.

Downloads last month
-
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for JoyFusionAI/dt-anima

Quantized
(47)
this model