diff --git a/litellm/__init__.py b/litellm/__init__.py
index ea5844320..9e7c26186 100644
--- a/litellm/__init__.py
+++ b/litellm/__init__.py
@@ -174,6 +174,7 @@ upperbound_key_generate_params: Optional[Dict] = None
 default_user_params: Optional[Dict] = None
 default_team_settings: Optional[List] = None
 max_user_budget: Optional[float] = None
+max_end_user_budget: Optional[float] = None
 #### RELIABILITY ####
 request_timeout: Optional[float] = 6000
 num_retries: Optional[int] = None  # per model endpoint
diff --git a/litellm/proxy/auth/auth_checks.py b/litellm/proxy/auth/auth_checks.py
index e9280fd31..794d393c1 100644
--- a/litellm/proxy/auth/auth_checks.py
+++ b/litellm/proxy/auth/auth_checks.py
@@ -69,7 +69,7 @@ def common_checks(
         end_user_budget = end_user_object.litellm_budget_table.max_budget
         if end_user_budget is not None and end_user_object.spend > end_user_budget:
             raise Exception(
-                f"End User={end_user_object.user_id} over budget. Spend={end_user_object.spend}, Budget={end_user_budget}"
+                f"ExceededBudget: End User={end_user_object.user_id} over budget. Spend={end_user_object.spend}, Budget={end_user_budget}"
             )
     # 5. [OPTIONAL] If 'enforce_user_param' enabled - did developer pass in 'user' param for openai endpoints
     if (
diff --git a/litellm/proxy/proxy_server.py b/litellm/proxy/proxy_server.py
index 156157976..e5d57c8a6 100644
--- a/litellm/proxy/proxy_server.py
+++ b/litellm/proxy/proxy_server.py
@@ -692,7 +692,6 @@ async def user_api_key_auth(
                 )
 
             # Check 2. If user_id for this token is in budget
-            ## Check 2.2 [OPTIONAL - checked only if litellm.max_user_budget is not None] If 'user' passed in /chat/completions is in budget
             if valid_token.user_id is not None:
                 user_id_list = [valid_token.user_id]
                 for id in user_id_list:
@@ -909,6 +908,13 @@ async def user_api_key_auth(
                 models=valid_token.team_models,
             )
 
+            _end_user_object = None
+            if "user" in request_data:
+                _id = "end_user_id:{}".format(request_data["user"])
+                _end_user_object = await user_api_key_cache.async_get_cache(key=_id)
+                if _end_user_object is not None:
+                    _end_user_object = LiteLLM_EndUserTable(**_end_user_object)
+
             global_proxy_spend = None
             if litellm.max_budget > 0:  # user set proxy max budget
                 # check cache
@@ -947,7 +953,7 @@ async def user_api_key_auth(
             _ = common_checks(
                 request_body=request_data,
                 team_object=_team_obj,
-                end_user_object=None,
+                end_user_object=_end_user_object,
                 general_settings=general_settings,
                 global_proxy_spend=global_proxy_spend,
                 route=route,
@@ -1617,7 +1623,7 @@ async def update_cache(
 
     async def _update_user_cache():
         ## UPDATE CACHE FOR USER ID + GLOBAL PROXY
-        user_ids = [user_id, litellm_proxy_budget_name, end_user_id]
+        user_ids = [user_id, litellm_proxy_budget_name]
         try:
             for _id in user_ids:
                 # Fetch the existing cost for the given user
@@ -1664,8 +1670,7 @@ async def update_cache(
             )
 
     async def _update_end_user_cache():
-        ## UPDATE CACHE FOR USER ID + GLOBAL PROXY
-        _id = end_user_id
+        _id = "end_user_id:{}".format(end_user_id)
         try:
             # Fetch the existing cost for the given user
             existing_spend_obj = await user_api_key_cache.async_get_cache(key=_id)
@@ -1673,14 +1678,14 @@ async def update_cache(
                 # if user does not exist in LiteLLM_UserTable, create a new user
                 existing_spend = 0
                 max_user_budget = None
-                if litellm.max_user_budget is not None:
-                    max_user_budget = litellm.max_user_budget
+                if litellm.max_end_user_budget is not None:
+                    max_end_user_budget = litellm.max_end_user_budget
                 existing_spend_obj = LiteLLM_EndUserTable(
                     user_id=_id,
                     spend=0,
                     blocked=False,
                     litellm_budget_table=LiteLLM_BudgetTable(
-                        max_budget=max_user_budget
+                        max_budget=max_end_user_budget
                     ),
                 )
             verbose_proxy_logger.debug(
diff --git a/litellm/proxy/utils.py b/litellm/proxy/utils.py
index f313a7985..f70d67aac 100644
--- a/litellm/proxy/utils.py
+++ b/litellm/proxy/utils.py
@@ -1941,9 +1941,9 @@ async def update_spend(
                             end_user_id,
                             response_cost,
                         ) in prisma_client.end_user_list_transactons.items():
-                            max_user_budget = None
-                            if litellm.max_user_budget is not None:
-                                max_user_budget = litellm.max_user_budget
+                            max_end_user_budget = None
+                            if litellm.max_end_user_budget is not None:
+                                max_end_user_budget = litellm.max_end_user_budget
                             new_user_obj = LiteLLM_EndUserTable(
                                 user_id=end_user_id, spend=response_cost, blocked=False
                             )
diff --git a/litellm/tests/test_key_generate_prisma.py b/litellm/tests/test_key_generate_prisma.py
index 103b344f5..29468c449 100644
--- a/litellm/tests/test_key_generate_prisma.py
+++ b/litellm/tests/test_key_generate_prisma.py
@@ -324,7 +324,7 @@ def test_call_with_end_user_over_budget(prisma_client):
 
     setattr(litellm.proxy.proxy_server, "prisma_client", prisma_client)
     setattr(litellm.proxy.proxy_server, "master_key", "sk-1234")
-    setattr(litellm, "max_user_budget", 0.00001)
+    setattr(litellm, "max_end_user_budget", 0.00001)
     try:
 
         async def test():
@@ -378,7 +378,9 @@ def test_call_with_end_user_over_budget(prisma_client):
                             "user_api_key_user_id": user,
                         },
                         "proxy_server_request": {
-                            "user": user,
+                            "body": {
+                                "user": user,
+                            }
                         },
                     },
                     "response_cost": 10,