diff --git a/litellm/proxy/_types.py b/litellm/proxy/_types.py
index f009390576a..ba8d9061d44 100644
--- a/litellm/proxy/_types.py
+++ b/litellm/proxy/_types.py
@@ -454,18 +454,23 @@ class LiteLLMRoutes(enum.Enum):
         "/responses",
         "/v1/responses",
         "/openai/v1/responses",
+        "/openai/responses",
         "/responses/{response_id}",
         "/v1/responses/{response_id}",
         "/openai/v1/responses/{response_id}",
+        "/openai/responses/{response_id}",
         "/responses/{response_id}/input_items",
         "/v1/responses/{response_id}/input_items",
         "/openai/v1/responses/{response_id}/input_items",
+        "/openai/responses/{response_id}/input_items",
         "/responses/{response_id}/cancel",
         "/v1/responses/{response_id}/cancel",
         "/openai/v1/responses/{response_id}/cancel",
+        "/openai/responses/{response_id}/cancel",
         "/responses/input_tokens",
         "/v1/responses/input_tokens",
         "/openai/v1/responses/input_tokens",
+        "/openai/responses/input_tokens",
         # vector stores
         "/vector_stores",
         "/v1/vector_stores",
diff --git a/litellm/proxy/agent_endpoints/auth/managed_authorization.py b/litellm/proxy/agent_endpoints/auth/managed_authorization.py
index 64e50cd74e7..0ea4060bcee 100644
--- a/litellm/proxy/agent_endpoints/auth/managed_authorization.py
+++ b/litellm/proxy/agent_endpoints/auth/managed_authorization.py
@@ -37,6 +37,7 @@ _MANAGED_MODEL_ROUTES: Final = frozenset(
 ) | frozenset(
     (
         "/openai/v1/responses",
+        "/openai/responses",
         "/v2/rerank",
         "/claude_code_gateway/v1/messages",
         "/claude_code_gateway/v1/messages/count_tokens",
diff --git a/litellm/proxy/common_utils/custom_openapi_spec.py b/litellm/proxy/common_utils/custom_openapi_spec.py
index c1d4442893a..f79dce9cff2 100644
--- a/litellm/proxy/common_utils/custom_openapi_spec.py
+++ b/litellm/proxy/common_utils/custom_openapi_spec.py
@@ -31,7 +31,7 @@ class CustomOpenAPISpec:
         "/openai/deployments/{model}/embeddings",
     ]
 
-    RESPONSES_API_PATHS = ["/v1/responses", "/responses", "/openai/v1/responses"]
+    RESPONSES_API_PATHS = ["/v1/responses", "/responses", "/openai/v1/responses", "/openai/responses"]
 
     @staticmethod
     def _as_object(node: JsonValue) -> JsonObject:
diff --git a/litellm/proxy/response_api_endpoints/endpoints.py b/litellm/proxy/response_api_endpoints/endpoints.py
index 7f9fcca369b..f6f72971b47 100644
--- a/litellm/proxy/response_api_endpoints/endpoints.py
+++ b/litellm/proxy/response_api_endpoints/endpoints.py
@@ -213,6 +213,12 @@ async def _resolve_cursor_model_variant_before_auth(request: Request) -> None:
     tags=["responses"],
     responses=RESPONSES_API_CREATE_RESPONSE_SCHEMAS,
 )
+@router.post(
+    "/openai/responses",
+    dependencies=[Depends(user_api_key_auth)],
+    tags=["responses"],
+    responses=RESPONSES_API_CREATE_RESPONSE_SCHEMAS,
+)
 async def responses_api(
     request: Request,
     fastapi_response: Response,
@@ -696,6 +702,12 @@ async def cursor_chat_completions(
     tags=["responses"],
     responses=RESPONSES_API_RESPONSE_SCHEMAS,
 )
+@router.get(
+    "/openai/responses/{response_id}",
+    dependencies=[Depends(user_api_key_auth)],
+    tags=["responses"],
+    responses=RESPONSES_API_RESPONSE_SCHEMAS,
+)
 async def get_response(
     response_id: str,
     request: Request,
@@ -812,6 +824,12 @@ async def get_response(
     tags=["responses"],
     responses=DELETE_RESPONSE_SCHEMAS,
 )
+@router.delete(
+    "/openai/responses/{response_id}",
+    dependencies=[Depends(user_api_key_auth)],
+    tags=["responses"],
+    responses=DELETE_RESPONSE_SCHEMAS,
+)
 async def delete_response(
     response_id: str,
     request: Request,
@@ -921,6 +939,12 @@ async def delete_response(
     tags=["responses"],
     responses=RESPONSE_ITEM_LIST_SCHEMAS,
 )
+@router.get(
+    "/openai/responses/{response_id}/input_items",
+    dependencies=[Depends(user_api_key_auth)],
+    tags=["responses"],
+    responses=RESPONSE_ITEM_LIST_SCHEMAS,
+)
 async def get_response_input_items(
     response_id: str,
     request: Request,
@@ -989,6 +1013,11 @@ async def get_response_input_items(
     dependencies=[Depends(user_api_key_auth)],
     tags=["responses"],
 )
+@router.post(
+    "/openai/responses/compact",
+    dependencies=[Depends(user_api_key_auth)],
+    tags=["responses"],
+)
 async def compact_response(
     request: Request,
     fastapi_response: Response,
@@ -1144,6 +1173,11 @@ def _responses_input_as_token_count_messages(
     dependencies=(_user_api_key_auth_dep,),
     tags=_RESPONSES_TAGS,
 )
+@router.post(
+    "/openai/responses/input_tokens",
+    dependencies=(_user_api_key_auth_dep,),
+    tags=_RESPONSES_TAGS,
+)
 async def responses_input_tokens(
     request: Request,
     token_counter: _TokenCounter = _token_counter_dep,
@@ -1217,6 +1251,11 @@ async def responses_input_tokens(
     dependencies=[Depends(user_api_key_auth)],
     tags=["responses"],
 )
+@router.post(
+    "/openai/responses/{response_id}/cancel",
+    dependencies=[Depends(user_api_key_auth)],
+    tags=["responses"],
+)
 async def cancel_response(
     response_id: str,
     request: Request,
diff --git a/litellm/proxy/spend_tracking/budget_reservation.py b/litellm/proxy/spend_tracking/budget_reservation.py
index 4c9df95cd8d..96f78c164cd 100644
--- a/litellm/proxy/spend_tracking/budget_reservation.py
+++ b/litellm/proxy/spend_tracking/budget_reservation.py
@@ -236,6 +236,7 @@ _UNBILLED_ROUTES: Final[frozenset[str]] = frozenset(
         "/responses/input_tokens",
         "/v1/responses/input_tokens",
         "/openai/v1/responses/input_tokens",
+        "/openai/responses/input_tokens",
     }
 )
 _TOKEN_COUNTING_SEGMENTS: Final[frozenset[str]] = frozenset({"count_tokens", "count-tokens"})
diff --git a/tests/proxy_unit_tests/test_openai_responses_routes.py b/tests/proxy_unit_tests/test_openai_responses_routes.py
new file mode 100644
index 00000000000..dc83a1fb257
--- /dev/null
+++ b/tests/proxy_unit_tests/test_openai_responses_routes.py
@@ -0,0 +1,27 @@
+from litellm.proxy._types import LiteLLMRoutes
+from litellm.proxy.agent_endpoints.auth.managed_authorization import _MANAGED_MODEL_ROUTES
+from litellm.proxy.common_utils.custom_openapi_spec import CustomOpenAPISpec
+from litellm.proxy.response_api_endpoints.endpoints import router
+from litellm.proxy.spend_tracking.budget_reservation import _UNBILLED_ROUTES
+
+
+def test_openai_responses_routes_registered():
+    """Verify that Azure-style /openai/responses endpoints are registered on the router."""
+    routes = {getattr(r, "path", None) for r in router.routes}
+
+    expected_routes = [
+        "/openai/responses",
+        "/openai/responses/{response_id}",
+        "/openai/responses/{response_id}/input_items",
+        "/openai/responses/compact",
+        "/openai/responses/input_tokens",
+        "/openai/responses/{response_id}/cancel",
+    ]
+
+    for expected in expected_routes:
+        assert expected in routes, f"Route {expected} missing from router.routes!"
+
+    assert "/openai/responses" in CustomOpenAPISpec.RESPONSES_API_PATHS
+    assert "/openai/responses" in _MANAGED_MODEL_ROUTES
+    assert "/openai/responses/input_tokens" in _UNBILLED_ROUTES
+    assert "/openai/responses" in LiteLLMRoutes.openai_routes.value