diff --git a/litellm/responses/streaming_iterator.py b/litellm/responses/streaming_iterator.py
index 2be9fdbc2a5..e42d845325a 100644
--- a/litellm/responses/streaming_iterator.py
+++ b/litellm/responses/streaming_iterator.py
@@ -1751,7 +1751,7 @@ def _estimate_usage_from_text(
     responses_api_request: Mapping[str, object],
     generated_text: str,
     *,
-    use_default_image_token_count: bool,
+    use_default_image_token_count: bool = False,
 ) -> ResponseAPIUsage:
     messages: Final = LiteLLMCompletionResponsesConfig.transform_responses_api_input_to_messages(  # pyright: ignore[reportUnknownMemberType]  # the transformer's signature is partially untyped
         input=request_input,  # pyright: ignore[reportArgumentType]  # the raw Responses API input is a str or ResponseInputParam list, matching the helper's declared union
diff --git a/tests/unit/responses/test_streaming_iterator.py b/tests/unit/responses/test_streaming_iterator.py
index 663e3b16a16..533b8d5267e 100644
--- a/tests/unit/responses/test_streaming_iterator.py
+++ b/tests/unit/responses/test_streaming_iterator.py
@@ -1206,7 +1206,7 @@ async def test_completed_event_survives_a_failing_usage_estimate():
     malformed_input: Final = [{"type": "message", "role": "user", "content": 42}]
     with pytest.raises(ValueError, match="Invalid content type"):
         _estimate_usage_from_text(
-            "gpt-4o-mini", malformed_input, {"input": malformed_input}, "hello world", use_default_image_token_count=False
+            "gpt-4o-mini", malformed_input, {"input": malformed_input}, "hello world"
         )
 
     response = _responses_api_response_without_usage()