diff --git a/client/ayon_core/hosts/houdini/plugins/publish/collect_instances.py b/client/ayon_core/hosts/houdini/plugins/publish/collect_instances.py
deleted file mode 100644
index 63537811cd..0000000000
--- a/client/ayon_core/hosts/houdini/plugins/publish/collect_instances.py
+++ /dev/null
@@ -1,100 +0,0 @@
-import hou
-
-import pyblish.api
-
-from ayon_core.pipeline import AYON_INSTANCE_ID, AVALON_INSTANCE_ID
-from ayon_core.hosts.houdini.api import lib
-
-
-class CollectInstances(pyblish.api.ContextPlugin):
-    """Gather instances by all node in out graph and pre-defined attributes
-
-    This collector takes into account folders that are associated with
-    an specific node and marked with a unique identifier;
-
-    Identifier:
-        id (str): "ayon.create.instance"
-
-    Specific node:
-        The specific node is important because it dictates in which way the
-        product is being exported.
-
-        alembic: will export Alembic file which supports cascading attributes
-                 like 'cbId' and 'path'
-        geometry: Can export a wide range of file types, default out
-
-    """
-
-    order = pyblish.api.CollectorOrder - 0.01
-    label = "Collect Instances"
-    hosts = ["houdini"]
-
-    def process(self, context):
-
-        nodes = hou.node("/out").children()
-        nodes += hou.node("/obj").children()
-
-        # Include instances in USD stage only when it exists so it
-        # remains backwards compatible with version before houdini 18
-        stage = hou.node("/stage")
-        if stage:
-            nodes += stage.recursiveGlob("*", filter=hou.nodeTypeFilter.Rop)
-
-        for node in nodes:
-
-            if not node.parm("id"):
-                continue
-
-            if node.evalParm("id") not in {
-                AYON_INSTANCE_ID, AVALON_INSTANCE_ID
-            }:
-                continue
-
-            # instance was created by new creator code, skip it as
-            # it is already collected.
-            if node.parm("creator_identifier"):
-                continue
-
-            has_family = node.evalParm("family")
-            assert has_family, "'%s' is missing 'family'" % node.name()
-
-            self.log.info(
-                "Processing legacy instance node {}".format(node.path())
-            )
-
-            data = lib.read(node)
-            # Check bypass state and reverse
-            if hasattr(node, "isBypassed"):
-                data.update({"active": not node.isBypassed()})
-
-            # temporarily translation of `active` to `publish` till issue has
-            # been resolved.
-            # https://github.com/pyblish/pyblish-base/issues/307
-            if "active" in data:
-                data["publish"] = data["active"]
-
-            # Create nice name if the instance has a frame range.
-            label = data.get("name", node.name())
-            label += " (%s)" % data["folderPath"]  # include folder in name
-
-            instance = context.create_instance(label)
-
-            # Include `families` using `family` data
-            product_type = data["family"]
-            data["productType"] = product_type
-            instance.data["families"] = [product_type]
-
-            instance[:] = [node]
-            instance.data["instance_node"] = node.path()
-            instance.data.update(data)
-
-        def sort_by_family(instance):
-            """Sort by family"""
-            return instance.data.get(
-                "families", instance.data.get("productType")
-            )
-
-        # Sort/grouped by family (preserving local index)
-        context[:] = sorted(context, key=sort_by_family)
-
-        return context