diff --git a/.circleci/config.yml b/.circleci/config.yml
index f939fed00..fc0bb5b98 100644
--- a/.circleci/config.yml
+++ b/.circleci/config.yml
@@ -89,7 +89,7 @@ jobs:
           name: Linting Testing
           command: |
             cd litellm
-            python -m pip install types-requests types-setuptools types-redis types-PyYAML
+            python -m pip install types-requests types-setuptools types-redis types-PyYAML types-cachetools
             if ! python -m mypy . --ignore-missing-imports; then
               echo "mypy detected errors"
               exit 1
diff --git a/litellm/caching.py b/litellm/caching.py
index c46dd3af8..78b4bd270 100644
--- a/litellm/caching.py
+++ b/litellm/caching.py
@@ -17,7 +17,6 @@ import traceback
 from datetime import timedelta
 from typing import Any, BinaryIO, List, Literal, Optional, Union
 
-from cachetools import LRUCache
 from openai._models import BaseModel as OpenAIObject
 
 import litellm
@@ -69,6 +68,8 @@ class InMemoryCache(BaseCache):
         """
         max_size_in_memory [int]: Maximum number of items in cache. done to prevent memory leaks. Use 200 items as a default
         """
+        from cachetools import LRUCache
+
         self.max_size_in_memory = max_size_in_memory or 200
         self.cache_dict: LRUCache = LRUCache(maxsize=self.max_size_in_memory)
         self.ttl_dict: LRUCache = LRUCache(maxsize=self.max_size_in_memory)