diff --git a/__init__.py b/__init__.py
index 8f2cfcf..20596a9 100644
--- a/__init__.py
+++ b/__init__.py
@@ -178,6 +178,7 @@ NODE_CONFIG = {
     "PathchSageAttentionKJ": {"class": PathchSageAttentionKJ, "name": "Patch Sage Attention KJ"},
     "LeapfusionHunyuanI2VPatcher": {"class": LeapfusionHunyuanI2V, "name": "Leapfusion Hunyuan I2V Patcher"},
     "VAELoaderKJ": {"class": VAELoaderKJ, "name": "VAELoader KJ"},
+    "ScheduledCFGGuidance": {"class": ScheduledCFGGuidance, "name": "Scheduled CFG Guidance"},
 
     #instance diffusion
     "CreateInstanceDiffusionTracking": {"class": CreateInstanceDiffusionTracking},
diff --git a/nodes/nodes.py b/nodes/nodes.py
index b625697..e312a5a 100644
--- a/nodes/nodes.py
+++ b/nodes/nodes.py
@@ -2399,4 +2399,60 @@ class VAELoaderKJ:
             vae_path = folder_paths.get_full_path_or_raise("vae", vae_name)
             sd = load_torch_file(vae_path)
         vae = VAE(sd=sd, device=device, dtype=dtype)
-        return (vae,)
\ No newline at end of file
+        return (vae,)
+
+from comfy.samplers import sampling_function, CFGGuider
+class Guider_ScheduledCFG(CFGGuider):
+
+    def set_cfg(self, cfg, start_percent, end_percent):
+        self.cfg = cfg
+        self.start_percent = start_percent
+        self.end_percent = end_percent
+
+    def predict_noise(self, x, timestep, model_options={}, seed=None):
+        steps = model_options["transformer_options"]["sample_sigmas"]
+        matched_step_index = (steps == timestep).nonzero()
+        if len(matched_step_index) > 0:
+            current_step_index = matched_step_index.item()
+        else:
+            for i in range(len(steps) - 1):
+                # walk from beginning of steps until crossing the timestep
+                if (steps[i] - timestep) * (steps[i + 1] - timestep) <= 0:
+                    current_step_index = i
+                    break
+            else:
+                current_step_index = 0
+        current_percent = current_step_index / (len(steps) - 1)
+
+        if self.start_percent <= current_percent <= self.end_percent:
+            if isinstance(self.cfg, list):
+                cfg = self.cfg[current_step_index]
+            else:
+                cfg = self.cfg
+            uncond = self.conds.get("negative", None)
+        else:
+            uncond = None
+            cfg = 1.0
+        return sampling_function(self.inner_model, x, timestep, uncond, self.conds.get("positive", None), cfg, model_options=model_options, seed=seed)            
+  
+class ScheduledCFGGuidance:
+    @classmethod
+    def INPUT_TYPES(s):
+        return {"required": { 
+                    "model": ("MODEL",),
+                    "positive": ("CONDITIONING", ),
+                    "negative": ("CONDITIONING", ),
+                    "cfg": ("FLOAT", {"default": 6.0, "min": 0.0, "max": 100.0, "step": 0.01}),
+                    "start_percent": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step":0.01}),
+                    "end_percent": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step":0.01}),
+                    },
+                }
+    RETURN_TYPES = ("GUIDER",)
+    FUNCTION = "get_guider"
+    CATEGORY = "KJNodes/experimental"
+
+    def get_guider(self, model, cfg, positive, negative, start_percent, end_percent):
+        guider = Guider_ScheduledCFG(model) 
+        guider.set_conds(positive, negative)
+        guider.set_cfg(cfg, start_percent, end_percent)
+        return (guider, )
\ No newline at end of file