diff --git a/test/oai/test_completion.py b/test/oai/test_completion.py
index f33533b7bf8..d8be8fcdfa4 100644
--- a/test/oai/test_completion.py
+++ b/test/oai/test_completion.py
@@ -231,7 +231,7 @@ def test_humaneval(num_samples=1):
         raise_on_ratelimit_or_timeout=False,
     )
     # assert response == -1
-    config_list = autogen.config_list_openai_aoai(KEY_LOC, exclude="aoai")
+    config_list = autogen.config_list_openai_aoai(KEY_LOC)
     # a minimal tuning example
     config, _ = autogen.Completion.tune(
         data=tune_data,
@@ -376,7 +376,7 @@ def test_math(num_samples=-1):
     ]
 
     autogen.Completion.set_cache(seed)
-    config_list = autogen.config_list_openai_aoai(KEY_LOC)[:2]
+    config_list = autogen.config_list_openai_aoai(KEY_LOC)
     vanilla_config = {
         "model": "text-davinci-003",
         "temperature": 1,
@@ -451,5 +451,5 @@ def my_average(results):
     # test_chatcompletion()
     # test_multi_model()
     # test_nocontext()
-    test_humaneval(1)
-    # test_math(1)
+    # test_humaneval(1)
+    test_math(1)