ComfyUI Node
Layout Inputs
Bring a Layout node a photo to trace and remix, and a vision-language model (Qwen3-VL) for its agent to draw boxes with. The model only loads when you ask the agent.
Layout Inputs
- clip
- image
- inputs
CategoryNynxz/Conditioning/Layout
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| clipopt | CLIP | A text encoder that can generate and see, e.g. Qwen3-VL-8B from CLIPLoader. Used by the Layout editor's agent only. | |
| imageopt | IMAGE | Shown in the Layout editor after a run, to trace or remix. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| inputs | NYNXZ_LAYOUT_INPUTS | Into Layout. |