diff --git a/litellm/llms/azure_ai/chat/transformation.py b/litellm/llms/azure_ai/chat/transformation.py
index bebf5e2ee5a..0ab386d544e 100644
--- a/litellm/llms/azure_ai/chat/transformation.py
+++ b/litellm/llms/azure_ai/chat/transformation.py
@@ -28,7 +28,7 @@ from litellm.secret_managers.main import get_secret_str
 from litellm.types.llms.openai import AllMessageValues
 from litellm.types.router import GenericLiteLLMParams
 from litellm.types.utils import ModelResponse, ProviderField
-from litellm.utils import _add_path_to_api_base, supports_tool_choice
+from litellm.utils import _add_path_to_api_base, supports_reasoning, supports_tool_choice
 
 if TYPE_CHECKING:
     from litellm.litellm_core_utils.tokenizer import Encoding as Tokenizer
@@ -111,7 +111,9 @@ class AzureAIStudioConfig(OpenAIConfig):
             drop_params=drop_params,
         )
 
-        if "max_completion_tokens" not in optional_params:
+        if "max_completion_tokens" not in optional_params or supports_reasoning(
+            model=model, custom_llm_provider="azure_ai"
+        ):
             return optional_params
 
         max_tokens: Final = optional_params["max_completion_tokens"]
diff --git a/tests/unit/llms/azure_ai/chat/test_azure_ai_transformation.py b/tests/unit/llms/azure_ai/chat/test_azure_ai_transformation.py
index 5a99a4ef95a..28568958569 100644
--- a/tests/unit/llms/azure_ai/chat/test_azure_ai_transformation.py
+++ b/tests/unit/llms/azure_ai/chat/test_azure_ai_transformation.py
@@ -169,6 +169,17 @@ def test_azure_ai_keeps_max_completion_tokens_for_gpt_5():
     assert "max_tokens" not in mapped_params
 
 
+def test_azure_ai_keeps_max_completion_tokens_for_reasoning_models(_local_model_cost_map: None) -> None:
+    mapped_params: Final = litellm.get_optional_params(
+        model="o3",
+        custom_llm_provider="azure_ai",
+        max_completion_tokens=256,
+    )
+
+    assert mapped_params["max_completion_tokens"] == 256
+    assert "max_tokens" not in mapped_params
+
+
 def test_azure_ai_keeps_params_without_max_completion_tokens():
     mapped_params: Final = litellm.get_optional_params(
         model="mistral-large-3",