diff --git a/tests/open_ai/test_chat_with_tool_call.py b/tests/open_ai/test_chat_with_tool_call.py
index 1481842..93eba77 100644
--- a/tests/open_ai/test_chat_with_tool_call.py
+++ b/tests/open_ai/test_chat_with_tool_call.py
@@ -119,6 +119,9 @@ async def test_chat_completion_with_tool_call_without_streaming(
         )
         trace = await trace_response.json()
 
+        for message in trace["messages"]:
+            message.pop("annotations", None)
+
         # Verify the trace messages
         expected_messages = history + [
             {
diff --git a/tests/open_ai/test_chat_without_tool_calls.py b/tests/open_ai/test_chat_without_tool_calls.py
index d9ae738..e8128a8 100644
--- a/tests/open_ai/test_chat_without_tool_calls.py
+++ b/tests/open_ai/test_chat_without_tool_calls.py
@@ -79,6 +79,9 @@ async def test_chat_completion(
         )
         trace = await trace_response.json()
 
+        for message in trace["messages"]:
+            message.pop("annotations", None)
+
         # Verify the trace messages
         assert trace["messages"] == [
             {
@@ -159,6 +162,9 @@ async def test_chat_completion_with_image(
             )
             trace = await trace_response.json()
 
+            for message in trace["messages"]:
+                message.pop("annotations", None)
+
             # Verify the trace messages
             assert trace["messages"] == [
                 {
@@ -227,6 +233,9 @@ async def test_chat_completion_with_invariant_key_in_openai_key_header(
         )
         trace = await trace_response.json()
 
+        for message in trace["messages"]:
+            message.pop("annotations", None)
+
         # Verify the trace messages
         assert trace["messages"] == [
             {