diff --git a/bigcode_eval/tasks/humanevalpack_openai.py b/bigcode_eval/tasks/humanevalpack_openai.py
index 742382747..958740a24 100644
--- a/bigcode_eval/tasks/humanevalpack_openai.py
+++ b/bigcode_eval/tasks/humanevalpack_openai.py
@@ -37,6 +37,7 @@ def get_prompt_base(doc, language="python"):
 
 import os
 import openai
+import litellm
 import jsonlines
 import termcolor
 
@@ -170,7 +171,8 @@ def __call__(self, prompt: str, n: int) -> str:
         ]
         while True:
             try:
-                response = openai.ChatCompletion.create(
+                # Equivalent to response = openai.ChatCompletion.create(
+                response = litellm.completion(
                     model=self._model,
                     messages=messages,
                     temperature=0.2,
@@ -199,8 +201,11 @@ def __call__(self, prompt: str, n: int) -> str:
         with jsonlines.open(f"completions_{LANGUAGE}_humanevalexplaindescribe.jsonl", "r") as f:
             descriptions = [line["raw_generation"][0] for line in f]
 
-    openai.organization = os.getenv("OPENAI_ORGANIZATION")
-    openai.api_key = os.getenv("OPENAI_API_KEY")
+    # Equivalent to:
+    # openai.organization = os.getenv("OPENAI_ORGANIZATION")
+    # openai.api_key = os.getenv("OPENAI_API_KEY")
+    litellm.organization = os.getenv("OPENAI_ORGANIZATION")
+    litellm.api_key = os.getenv("OPENAI_API_KEY")
 
     samples = [s for s in load_dataset("bigcode/humanevalpack", LANGUAGE)["test"]]
 
diff --git a/setup.py b/setup.py
index cf80c7d24..663d33e7d 100644
--- a/setup.py
+++ b/setup.py
@@ -12,6 +12,7 @@
     "matplotlib==3.5.2",
     "numpy==1.21.6",
     "openai==0.23.0",
+    "litellm==0.13.2",
     "pandas==1.3.5",
     "pandas-datareader==0.10.0",
     "pathlib==1.0.1",