diff --git a/litellm/litellm_core_utils/streaming_handler.py b/litellm/litellm_core_utils/streaming_handler.py
index 9deac7923b6..a19be338edb 100644
--- a/litellm/litellm_core_utils/streaming_handler.py
+++ b/litellm/litellm_core_utils/streaming_handler.py
@@ -2010,6 +2010,8 @@ class CustomStreamWrapper:
                 if self.stream_options is None:  # add usage as hidden param
                     usage = calculate_total_usage(chunks=self.chunks)
                     processed_chunk.hidden_params["usage"] = usage
+                elif self._continuous_usage_stats:
+                    processed_chunk.usage = calculate_total_usage(chunks=self.chunks)
                 ## LOGGING
                 executor.submit(
                     self.run_success_logging_and_cache_storage,
@@ -2289,6 +2291,8 @@ class CustomStreamWrapper:
                     dict[str, object], getattr(processed_chunk, HIDDEN_PARAMS_ATTR)
                 )
                 processed_chunk_hidden_params["usage"] = usage
+            elif self._continuous_usage_stats:
+                processed_chunk.usage = calculate_total_usage(chunks=self.chunks)
             # see sync __next__'s sibling branch: deliberately do NOT restore
             # here - this chunk is still this call's own data, and restoring
             # before returning it would corrupt the caller's own log
diff --git a/tests/unit/litellm_core_utils/test_streaming_handler.py b/tests/unit/litellm_core_utils/test_streaming_handler.py
index db0f879a08d..21193041c16 100644
--- a/tests/unit/litellm_core_utils/test_streaming_handler.py
+++ b/tests/unit/litellm_core_utils/test_streaming_handler.py
@@ -6774,7 +6774,10 @@ async def test_continuous_usage_stats_finishes_before_the_final_usage_chunk(sync
     assert finish_indexes[0] < len(emitted) - 1
     terminal_chunk: Final = emitted[-1]
     assert not (terminal_chunk.choices and terminal_chunk.choices[0].finish_reason)
-    assert terminal_chunk.usage.total_tokens == final_usage["total_tokens"]
+    assert [chunk.usage.total_tokens for chunk in emitted] == [
+        content_usage["total_tokens"],
+        *[final_usage["total_tokens"]] * 3,
+    ]
 
 
 @pytest.mark.parametrize("sync_mode", [True, False])