diff --git a/litellm/cost_calculator.py b/litellm/cost_calculator.py
index af7dd078107..870b97530fa 100644
--- a/litellm/cost_calculator.py
+++ b/litellm/cost_calculator.py
@@ -587,6 +587,24 @@ def _model_contains_known_llm_provider(model: str) -> bool:
     return _provider_prefix in LlmProvidersSet
 
 
+def _get_response_model(completion_response: Any) -> Optional[str]:
+    """
+    Extract the model name from a completion response object.
+
+    Used as a fallback for cost calculation when the input model name
+    doesn't exist in model_cost (e.g., Azure Model Router).
+    """
+    if completion_response is None:
+        return None
+
+    if isinstance(completion_response, BaseModel):
+        return getattr(completion_response, "model", None)
+    elif isinstance(completion_response, dict):
+        return completion_response.get("model", None)
+
+    return None
+
+
 def _get_usage_object(
     completion_response: Any,
 ) -> Optional[Usage]:
@@ -933,9 +951,8 @@ def completion_cost(  # noqa: PLR0915
             router_model_id=router_model_id,
         )
 
-        potential_model_names = [selected_model]
-        if model is not None:
-            potential_model_names.append(model)
+        potential_model_names = [selected_model, _get_response_model(completion_response)]
+
 
         for idx, model in enumerate(potential_model_names):
             try: