ComfyUI Node
LZ MOSS Image Loader
A ComfyUI node in MyCustomNodes/Loaders with 6 inputs and 10 outputs.
LZ MOSS Image Loader
- MODEL
- CLIP
- VAE
- positive
- negative
- positive_text
- negative_text
- lz_pipe
- model_name
- model_hash
◄diffusion_model▾►
◄weight_dtypedefault►
◄text_encoder▾►
◄vae▾►
◄positive►
◄negativensfw, worst quality, bad anatomy, watermark, jpeg artifact►
CategoryMyCustomNodes/Loaders
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| diffusion_model | COMBO | 0 options: | |
| weight_dtype | COMBO | default | Model weight precision. Use fp8 variants to reduce VRAM (matches UNETLoader options). |
| text_encoder | COMBO | Use the Qwen3.5 9B text encoder (e.g. qwen3.5_9b_w4a8_gptq.safetensors). | |
| vae | COMBO | Use the FLUX.2 VAE (e.g. flux2-vae_small.safetensors). | |
| positive | STRING | — | |
| negative | STRING | nsfw, worst quality, bad anatomy, watermark, jpeg artifact | — |
Outputs (10)
| Name | Type | Description |
|---|---|---|
| MODEL | MODEL | — |
| CLIP | CLIP | — |
| VAE | VAE | — |
| positive | CONDITIONING | — |
| negative | CONDITIONING | — |
| positive_text | STRING | — |
| negative_text | STRING | — |
| lz_pipe | LZ_PIPE | — |
| model_name | STRING | — |
| model_hash | STRING | — |