Skip to content

The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface. The fastest local inference engine in the world.

Searching "kitchen" now returns the attention node. (#16632)

comfyanonymous2139131e880c
diff --git a/comfy_extras/nodes_model_advanced.py b/comfy_extras/nodes_model_advanced.py
index 97adc8cd8..c1917d3d9 100644
--- a/comfy_extras/nodes_model_advanced.py
+++ b/comfy_extras/nodes_model_advanced.py
@@ -380,6 +380,7 @@ class ModelAttentionBackend(io.ComfyNode):
return io.Schema(
node_id="ModelAttentionBackend",
display_name="Model Attention Backend",
+ search_aliases=["comfy kitchen", "attention", "pytorch", "kitchen attention"],
category="model/patch",
is_experimental=True,
description="Selects the dense attention implementation for the model. When used with Block Sparse Attention, this backend is used whenever sparse attention is inactive or unsupported.",