ComfyUI Node
EasyLoader (Full) - Anima
An Anima-only EasyLoader. By default it loads the diffusion model, Qwen3 0.6B text encoder, and VAE separately. An optional verified Anima checkpoint can override the model and any components it includes.
EasyLoader (Full) - Anima
- pipe
- model
- vae
- clip
- latent
- model_name
- vae_name
◄checkPointNone (use separate model / CLIP / VAE)►
◄model_name▾►
◄weight_dtypedefault►
◄clip_name▾►
◄clip_devicedefault►
◄vae_name▾►
◄resolution1024 x 1024►
◄empty_latent_width1024►
◄empty_latent_height1024►
◄batch_size1►
CategoryEasyUse-Anima/Loaders
Inputs (10)
| Name | Type | Default | Description |
|---|---|---|---|
| checkPoint | COMBO | None (use separate model / CLIP / VAE) | Anima checkpoint from models/checkpoints. Its MODEL is used, included CLIP/VAE take priority, and missing CLIP/VAE fall back to the saved selections below. |
| model_name | COMBO | Anima diffusion model from models/diffusion_models; for example anima-base-v1.0.safetensors. | |
| weight_dtype | COMBO | default | 4 options: default, fp8_e4m3fn, fp8_e4m3fn_fast, fp8_e5m2 |
| clip_name | COMBO | Anima Qwen3 0.6B text encoder from models/text_encoders; normally qwen_3_06b_base.safetensors. | |
| clip_device | COMBO | default | 2 options: default, cpu |
| vae_name | COMBO | Anima requires a 16-channel Qwen Image VAE; normally qwen_image_vae.safetensors. Incompatible SD VAEs are rejected. | |
| resolution | COMBO | 1024 x 1024 | 31 options: width x height (custom), 512 x 512, 512 x 768, 576 x 1024, 768 x 512, 768 x 768, +25 |
| empty_latent_width | INT | 102464–16384 | — |
| empty_latent_height | INT | 102464–16384 | — |
| batch_size | INT | 11–4096 | — |
Outputs (7)
| Name | Type | Description |
|---|---|---|
| pipe | PIPE_LINE | — |
| model | MODEL | — |
| vae | VAE | — |
| clip | CLIP | — |
| latent | LATENT | — |
| model_name | — | |
| vae_name | Baked VAE | — |