diff --git a/deepinfer/environ.py b/deepinfer/environ.py
index cbae526..9c6fc9e 100644
--- a/deepinfer/environ.py
+++ b/deepinfer/environ.py
@@ -114,6 +114,8 @@ class envs:
     DEEPINFER_NEURON_GRAPH_MODE = EnvVar.str(default="regions")
     # Neuron only: cut the region graphs further at the stage boundaries marked by neuron_ops.regions.split().
     DEEPINFER_NEURON_REGION_SPLIT = EnvVar.bool(default=False)
+    # Neuron only: one region graph per decoder layer (attention half + MoE) instead of two.
+    DEEPINFER_NEURON_REGION_LAYER = EnvVar.bool(default=False)
     # Neuron only: split the region graphs of prefill chunks (steps of more than one row) regardless.
     DEEPINFER_NEURON_REGION_SPLIT_PREFILL = EnvVar.bool(default=True)
     # Neuron only: synchronize around every region graph call and log per-kind times each decode step (profiling).
diff --git a/deepinfer/model_executor/models/decoder_layer.py b/deepinfer/model_executor/models/decoder_layer.py
index 3556839..e85c368 100644
--- a/deepinfer/model_executor/models/decoder_layer.py
+++ b/deepinfer/model_executor/models/decoder_layer.py
@@ -115,6 +115,12 @@ class DeepseekDecoderLayer(HCLayerMixin):
         # A region graph takes the fetched Engram rows as a tensor input rather than the per-step fetch hook.
         engram_input = embed_hook() if use_regions and embed_hook is not None else embed_hook
 
+        if use_regions and envs.DEEPINFER_NEURON_REGION_LAYER:
+            # One region graph for the whole layer (attention half + MoE): half the NEFF launches a step.
+            return neuron_regions.entry(self, "_layer_body")(
+                hidden_states, kv_cache, input_metadata, residual, engram_input, cross_layer_runtime, dflash_input
+            )
+
         second_layer, second_last_layer, same_layers = get_enable_timer_layer(self.layer_idx)
         attn_timer_scale = len(same_layers)
         with (
@@ -159,6 +165,14 @@ class DeepseekDecoderLayer(HCLayerMixin):
 
         return hidden_states, residual
 
+    def _layer_body(self, hidden_states, kv_cache, input_metadata, residual, engram_input, cross_layer_runtime, dflash_input):  # noqa: ANN001, ANN202
+        """The layer as one Neuron region graph (DEEPINFER_NEURON_REGION_LAYER)."""
+        hidden_states, hidden_states_bf16, residual = self._attn_half(
+            hidden_states, kv_cache, input_metadata, residual, engram_input, cross_layer_runtime, dflash_input
+        )
+        neuron_regions.split()
+        return self.mlp(hidden_states, hidden_states_bf16, input_metadata), residual
+
     def _attn_half(
         self,
         hidden_states: torch.Tensor,