diff --git a/src/diffusers/models/modeling_utils.py b/src/diffusers/models/modeling_utils.py
index 1e9e28471d89..a0f96390062a 100644
--- a/src/diffusers/models/modeling_utils.py
+++ b/src/diffusers/models/modeling_utils.py
@@ -266,6 +266,7 @@ class ModelMixin(torch.nn.Module, PushToHubMixin):
     _keep_in_fp32_modules = None
     _skip_layerwise_casting_patterns = None
     _supports_group_offloading = True
+    _regions_for_compile = []
 
     def __init__(self):
         super().__init__()
@@ -1402,6 +1403,44 @@ def float(self, *args):
         else:
             return super().float(*args)
 
+    @wraps(torch.nn.Module.compile)
+    def compile(self, use_regional_compile: bool = True, *args, **kwargs):
+        """ """
+        if use_regional_compile:
+            regions_for_compile = getattr(self, "_regions_for_compile", None)
+
+            if not regions_for_compile:
+                logger.warning(
+                    "_regions_for_compile attribute is empty. Using _no_split_modules to find compile regions."
+                )
+
+                regions_for_compile = getattr(self, "_no_split_modules", None)
+
+            if not regions_for_compile:
+                logger.warning(
+                    "Both _regions_for_compile and _no_split_modules attribute are empty. "
+                    "Set _regions_for_compile for the model to benefit from regional compilation. "
+                    "Falling back to full model compilation, which could have high first iteration "
+                    "latency."
+                )
+                super().compile(*args, **kwargs)
+
+            has_compiled_region = False
+            for submod in self.modules():
+                if submod.__class__.__name__ in regions_for_compile:
+                    has_compiled_region = True
+                    submod.compile(*args, **kwargs)
+
+            if not has_compiled_region:
+                raise ValueError(
+                    f"Regional compilation failed because {regions_for_compile} classes are not found in the model. "
+                    "Either set them correctly, or set `use_regional_compile` to False while calling copmile, e.g. "
+                    "pipe.transformer.compile(use_regional_compile=False) to fallback to full model compilation, "
+                    "which could have high iteration latency."
+                )
+        else:
+            super().compile(*args, **kwargs)
+
     @classmethod
     def _load_pretrained_model(
         cls,
diff --git a/src/diffusers/models/transformers/transformer_flux.py b/src/diffusers/models/transformers/transformer_flux.py
index 541576b13b78..a0d2155cc025 100644
--- a/src/diffusers/models/transformers/transformer_flux.py
+++ b/src/diffusers/models/transformers/transformer_flux.py
@@ -227,6 +227,7 @@ class FluxTransformer2DModel(
     _supports_gradient_checkpointing = True
     _no_split_modules = ["FluxTransformerBlock", "FluxSingleTransformerBlock"]
     _skip_layerwise_casting_patterns = ["pos_embed", "norm"]
+    _regions_for_compile = _no_split_modules
 
     @register_to_config
     def __init__(
diff --git a/src/diffusers/models/transformers/transformer_wan.py b/src/diffusers/models/transformers/transformer_wan.py
index baa0ede4184e..fb693804c6fe 100644
--- a/src/diffusers/models/transformers/transformer_wan.py
+++ b/src/diffusers/models/transformers/transformer_wan.py
@@ -345,6 +345,7 @@ class WanTransformer3DModel(ModelMixin, ConfigMixin, PeftAdapterMixin, FromOrigi
     _no_split_modules = ["WanTransformerBlock"]
     _keep_in_fp32_modules = ["time_embedder", "scale_shift_table", "norm1", "norm2", "norm3"]
     _keys_to_ignore_on_load_unexpected = ["norm_added_q"]
+    _regions_for_compile = _no_split_modules
 
     @register_to_config
     def __init__(