exposing a litellm.max budget

litellm/__init__.py

@@ -36,7 +36,8 @@ caching = False # deprecated son
 caching_with_models = False  # if you want the caching key to be model + prompt # deprecated soon
 cache: Optional[Cache] = None # cache object
 model_alias_map: Dict[str, str] = {}
-
+max_budget = None # set the max budget across all providers
+_current_cost = 0 # private variable, used if max budget is set 
 #############################################
 
 def get_model_cost_map():

litellm/exceptions.py

@@ -108,3 +108,10 @@ class OpenAIError(OpenAIError):  # type: ignore
             code=original_exception.code,
         )
         self.llm_provider = "openai"
+
+class BudgetExceededError(Exception):
+    def __init__(self, current_cost, max_budget):
+        self.current_cost = current_cost
+        self.max_budget = max_budget
+        message = f"Budget has been exceeded! Current cost: {current_cost}, Max budget: {max_budget}"
+        super().__init__(message)

litellm/tests/test_budget_manager.py

@@ -10,7 +10,7 @@ sys.path.insert(
 )  # Adds the parent directory to the system path
 import litellm 
 litellm.set_verbose = True
-from litellm import BudgetManager, completion 
+from litellm import completion, BudgetManager
 
 budget_manager = BudgetManager(project_name="test_project", client_type="hosted")
 

litellm/tests/test_litellm_max_budget.py

@@ -0,0 +1,21 @@
+#### What this tests ####
+#    This tests calling litellm.max_budget by making back-to-back gpt-4 calls
+# commenting out this test for circle ci, as it causes other tests to fail, since litellm.max_budget would impact other litellm imports
+# import sys, os, json
+# import traceback
+# import pytest 
+
+# sys.path.insert(
+#     0, os.path.abspath("../..")
+# )  # Adds the parent directory to the system path
+# import litellm 
+# litellm.set_verbose = True
+# from litellm import completion
+
+# litellm.max_budget = 0.001 # sets a max budget of $0.001
+
+# messages = [{"role": "user", "content": "Hey, how's it going"}]
+# completion(model="gpt-4", messages=messages)
+# completion(model="gpt-4", messages=messages)
+# print(litellm._current_cost)
+

litellm/utils.py

@@ -31,7 +31,8 @@ from .exceptions import (
     ContextWindowExceededError,
     Timeout,
     APIConnectionError,
-    APIError
+    APIError,
+    BudgetExceededError
 )
 from typing import List, Dict, Union, Optional
 from .caching import Cache
@@ -542,6 +543,12 @@ def client(original_function):
         try:
             logging_obj = function_setup(start_time, *args, **kwargs)
             kwargs["litellm_logging_obj"] = logging_obj
+
+            # [OPTIONAL] CHECK BUDGET 
+            if litellm.max_budget:
+                if litellm._current_cost > litellm.max_budget:
+                    raise BudgetExceededError(current_cost=litellm._current_cost, max_budget=litellm.max_budget)
+
             # [OPTIONAL] CHECK CACHE
             # remove this after deprecating litellm.caching
             if (litellm.caching or litellm.caching_with_models) and litellm.cache is None:
@@ -567,6 +574,10 @@ def client(original_function):
             if litellm.caching or litellm.caching_with_models or litellm.cache != None: # user init a cache object
                 litellm.cache.add_cache(result, *args, **kwargs)
             
+            # [OPTIONAL] UPDATE BUDGET
+            if litellm.max_budget: 
+                litellm._current_cost += litellm.completion_cost(completion_response=result)
+            
             # [OPTIONAL] Return LiteLLM call_id
             if litellm.use_client == True:
                 result['litellm_call_id'] = litellm_call_id

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "litellm"
-version = "0.1.628"
+version = "0.1.629"
 description = "Library to easily interface with LLM API providers"
 authors = ["BerriAI"]
 license = "MIT License"