diff --git a/strix/llm/request_log.py b/strix/llm/request_log.py
index eda0d222e..89b421689 100644
--- a/strix/llm/request_log.py
+++ b/strix/llm/request_log.py
@@ -779,17 +779,24 @@ def _litellm_success(
     cost = _float_or_none(kwargs.get("response_cost"))
     if cost is None:
         cost = _float_or_none(hidden.get("response_cost"))
+    finish_reason = _litellm_finish_reason(response)
+    provider_filtered = finish_reason == "content_filter"
     return replace(
         base,
+        outcome="error" if provider_filtered else base.outcome,
         status_code=200,
         response_id=_str_or_none(getattr(response, "id", None)),
+        error_type="ProviderContentFilter" if provider_filtered else None,
+        error_message=(
+            "Response withheld by the provider's content filter." if provider_filtered else None
+        ),
         input_tokens=usage.get("input_tokens"),
         output_tokens=usage.get("output_tokens"),
         cached_input_tokens=usage.get("cached_input_tokens"),
         total_tokens=usage.get("total_tokens"),
         cost_usd=cost,
         response_bytes=json_size(response),
-        finish_reason=_litellm_finish_reason(response),
+        finish_reason=finish_reason,
         details=merge_details(
             ("request", _litellm_request_details(kwargs)),
             ("response", _litellm_response_details(response)),
diff --git a/tests/test_llm_request_log.py b/tests/test_llm_request_log.py
index 41bbc8843..50f0305bc 100644
--- a/tests/test_llm_request_log.py
+++ b/tests/test_llm_request_log.py
@@ -216,6 +216,25 @@ def test_litellm_success_falls_back_to_response_hidden_headers() -> None:
     assert event.provider_request_id == "req_hidden_1"
 
 
+def test_litellm_content_filter_success_callback_is_logged_as_provider_error() -> None:
+    class _Choice:
+        finish_reason = "content_filter"
+
+    class _Response(_FakeResponse):
+        choices: ClassVar[list[Any]] = [_Choice()]
+
+    event = request_log.event_from_litellm(
+        _anthropic_kwargs(None), _Response(), None, None, outcome="success"
+    )
+
+    assert event.outcome == "error"
+    assert event.status_code == 200
+    assert event.finish_reason == "content_filter"
+    assert event.error_type == "ProviderContentFilter"
+    assert event.error_message == "Response withheld by the provider's content filter."
+    assert (event.input_tokens, event.output_tokens, event.total_tokens) == (120, 30, 150)
+
+
 def test_litellm_anthropic_block_prefers_header_id_over_body_id() -> None:
     exc = _anthropic_error(400, ANTHROPIC_BLOCK_BODY, {"request-id": "req_HeaderWins01"})
     kwargs = _anthropic_kwargs(exc, headers={"request-id": "req_HeaderWins01"})