diff --git a/litellm/completion_extras/litellm_responses_transformation/transformation.py b/litellm/completion_extras/litellm_responses_transformation/transformation.py
index 5d5bcc6e1e9..71d3f1e900e 100644
--- a/litellm/completion_extras/litellm_responses_transformation/transformation.py
+++ b/litellm/completion_extras/litellm_responses_transformation/transformation.py
@@ -1334,6 +1334,7 @@ class OpenAiResponsesToChatCompletionStreamIterator(BaseModelResponseIterator):
     ):
         super().__init__(streaming_response, sync_stream, json_mode)
         self._chat_completion_id: str | None = None
+        self._served_service_tier: str | None = None
         self._tool_call_index_map: dict[int, int] = {}  # mutable-ok: per-stream accumulator state
 
     def _handle_string_chunk(
@@ -1645,12 +1646,28 @@ class OpenAiResponsesToChatCompletionStreamIterator(BaseModelResponseIterator):
             ModelResponseStream: OpenAI-formatted streaming chunk
         """
         verbose_logger.debug("Chat provider: transform_streaming_response called with chunk: %s", chunk)
-        return self._with_stream_scoped_id(
-            OpenAiResponsesToChatCompletionStreamIterator.translate_responses_chunk_to_openai_stream(
-                chunk, tool_call_index_map=self._tool_call_index_map
+        self._remember_served_service_tier(chunk)
+        return self._with_served_service_tier(
+            self._with_stream_scoped_id(
+                OpenAiResponsesToChatCompletionStreamIterator.translate_responses_chunk_to_openai_stream(
+                    chunk, tool_call_index_map=self._tool_call_index_map
+                )
             )
         )
 
+    def _remember_served_service_tier(self, chunk: dict[str, object]) -> None:
+        response_payload: Final = chunk.get("response")
+        if not isinstance(response_payload, dict):
+            return
+        served_tier: Final = response_payload.get("service_tier")
+        if isinstance(served_tier, str) and served_tier:
+            self._served_service_tier = served_tier
+
+    def _with_served_service_tier(self, chunk: "ModelResponseStream") -> "ModelResponseStream":
+        if self._served_service_tier is not None and chunk.model_dump().get("service_tier") is None:
+            setattr(chunk, "service_tier", self._served_service_tier)  # noqa: B010  # pydantic extra, not a declared field
+        return chunk
+
     def _with_stream_scoped_id(self, chunk: "ModelResponseStream") -> "ModelResponseStream":
         if self._chat_completion_id is None:
             self._chat_completion_id = chunk.id
diff --git a/tests/test_litellm/completion_extras/litellm_responses_transformation/test_completion_extras_litellm_responses_transformation_transformation.py b/tests/test_litellm/completion_extras/litellm_responses_transformation/test_completion_extras_litellm_responses_transformation_transformation.py
index 1d0e1be7c0f..282b84104a6 100644
--- a/tests/test_litellm/completion_extras/litellm_responses_transformation/test_completion_extras_litellm_responses_transformation_transformation.py
+++ b/tests/test_litellm/completion_extras/litellm_responses_transformation/test_completion_extras_litellm_responses_transformation_transformation.py
@@ -4369,3 +4369,22 @@ def test_response_completed_carries_the_served_service_tier():
     )
 
     assert result.model_dump()["service_tier"] == "default"
+
+
+def test_every_bridged_chunk_after_response_created_carries_the_served_service_tier():
+    from litellm.completion_extras.litellm_responses_transformation.transformation import (
+        OpenAiResponsesToChatCompletionStreamIterator,
+    )
+
+    iterator = OpenAiResponsesToChatCompletionStreamIterator(streaming_response=None, sync_stream=True)
+    events = [
+        {"type": "response.created", "response": {"id": "resp_1", "status": "in_progress", "service_tier": "default"}},
+        {"type": "response.output_item.added", "output_index": 0, "item": {"type": "message"}},
+        {"type": "response.output_text.delta", "output_index": 0, "delta": "Hi"},
+        {"type": "response.output_item.done", "output_index": 0, "item": {"type": "message"}},
+        {"type": "response.completed", "response": {"id": "resp_1", "status": "completed", "output": []}},
+    ]
+
+    relayed = [iterator.chunk_parser(event).model_dump().get("service_tier") for event in events]
+
+    assert relayed == ["default"] * len(events), relayed