diff --git a/backends/exllamav3/model.py b/backends/exllamav3/model.py
index 0d1780bc..b2d323bd 100644
--- a/backends/exllamav3/model.py
+++ b/backends/exllamav3/model.py
@@ -917,6 +917,9 @@ async def generate_gen(
         # Apply temperature last to builder
         if params.temperature_last:
             sampler_builder.temperature(params.temperature)
+        
+        # Apply adaptive P
+        sampler_builder.adaptive_p(params.adaptive_target, params.adaptive_decay)
 
         # Build the sampler
         # Set greedy if temperature is 0
diff --git a/backends/exllamav3/sampler.py b/backends/exllamav3/sampler.py
index 7b08a9b1..688cd183 100644
--- a/backends/exllamav3/sampler.py
+++ b/backends/exllamav3/sampler.py
@@ -11,6 +11,7 @@
     SS_TopP,
     SS_Sample,
     SS_Base,
+    SS_AdaptiveP,
 )
 
 
@@ -21,6 +22,7 @@ class ExllamaV3SamplerBuilder:
     """
 
     stack: List[SS_Base] = field(default_factory=list)
+    has_adaptive: bool = False
 
     def penalties(self, rep_p, freq_p, pres_p, penalty_range, rep_decay):
         self.stack += [
@@ -40,6 +42,11 @@ def top_p(self, top_p):
     def min_p(self, min_p):
         self.stack.append(SS_MinP(min_p))
 
+    def adaptive_p(self, adaptive_target, adaptive_decay):
+        self.stack.append(SS_AdaptiveP(adaptive_target, adaptive_decay))
+        if adaptive_target != 1.0:
+            self.has_adaptive = True
+
     def greedy(self):
         self.stack.append(SS_Argmax())
 
@@ -50,5 +57,6 @@ def build(self, greedy):
         if greedy:
             return CustomSampler([SS_Argmax()])
         else:
-            self.stack.append(SS_Sample())
+            if not self.has_adaptive:
+                self.stack.append(SS_Sample())
             return CustomSampler(self.stack)
diff --git a/common/sampling.py b/common/sampling.py
index 49be5b99..7333ee59 100644
--- a/common/sampling.py
+++ b/common/sampling.py
@@ -104,6 +104,13 @@ class BaseSamplerRequest(BaseModel):
     min_p: Optional[float] = Field(
         default_factory=lambda: get_default_sampler_value("min_p", 0.0)
     )
+    adaptive_target: Optional[float] = Field(
+        default_factory=lambda: get_default_sampler_value("adaptive_target", 0.0)
+    )
+
+    adaptive_decay: Optional[float] = Field(
+        default_factory=lambda: get_default_sampler_value("adaptive_decay", 0.0)
+    )
 
     tfs: Optional[float] = Field(
         default_factory=lambda: get_default_sampler_value("tfs", 1.0),