diff --git a/comfy/clip_vision.py b/comfy/clip_vision.py
index 20dc3345d..4f7a93290 100644
--- a/comfy/clip_vision.py
+++ b/comfy/clip_vision.py
@@ -36,6 +36,8 @@ class ClipVisionModel():
 
         self.image_size = config.get("image_size", 224)
         self.load_device = comfy.model_management.text_encoder_device()
+        # Force CPU for clip vision to work on IPEX
+        self.load_device = 'cpu' 
         offload_device = comfy.model_management.text_encoder_offload_device()
         self.dtype = comfy.model_management.text_encoder_dtype(self.load_device)
         self.model = comfy.clip_model.CLIPVisionModelProjection(config, self.dtype, offload_device, comfy.ops.manual_cast)
diff --git a/comfy/model_management.py b/comfy/model_management.py
index b3b8a394e..3bd13102a 100644
--- a/comfy/model_management.py
+++ b/comfy/model_management.py
@@ -540,6 +540,8 @@ def load_models_gpu(models, memory_required=0, force_patch_weights=False, minimu
 
 
 def load_model_gpu(model):
+    if model.load_device == 'cpu':
+        return 'cpu'
     return load_models_gpu([model])
 
 def loaded_models(only_currently_used=False):