This page is only for IMAGE_GENERATION. It loads enabled image-generation nodes dynamically from /api/nodes.
Reference images are optional. The page allows drag-drop or browse, and the backend remains the authority
for specialty, image count, and workflow compatibility.
Frontend sends what the user selected. Backend decides if it is valid.
Example specialties may expect different image counts. For example: SPRITE_ANIMATION may expect 2 images, POSE_ESTIMATION may expect 1 image, while GENERAL may work with 0 images.
Use the node info area below as guidance, but if something is invalid, the router should return the actual error.
Provider list comes from nodes where supportsImageGeneration = true.
Node info will appear here...
You can drag an image from a file folder or another web page into the drop area.
Image slot 2 automatically opens after slot 1 gets an image.
Image Slot 1
Images[0]
Primary reference image. Backend will decide how the selected node uses it.
Drop first image hereOr click to browse. Works for files dragged from folder or images dragged from another browser page.
Image Slot 2
Images[1]
Secondary reference image. Use when your selected workflow expects another image.
Drop second image hereThis placeholder opens after slot 1 gets an image, but you can still leave it empty and let backend decide.