diff --git a/strix/core/inputs.py b/strix/core/inputs.py
index 1326d8ac..7026cfe1 100644
--- a/strix/core/inputs.py
+++ b/strix/core/inputs.py
@@ -201,9 +201,12 @@ def make_model_settings(
     request_timeout: float | None = None,
     prompt_cache: bool = True,
     extra_headers: dict[str, str] | None = None,
+    has_tools: bool = True,
 ) -> ModelSettings:
+    # Azure OpenAI rejects any request that carries ``parallel_tool_calls``
+    # without ``tools``, so tool-less requests must omit it entirely.
     model_settings = ModelSettings(
-        parallel_tool_calls=False,
+        parallel_tool_calls=False if has_tools else None,
         retry=DEFAULT_MODEL_RETRY,
         include_usage=True,
         extra_args=request_timeout_extra_args(request_timeout),
diff --git a/strix/interface/main.py b/strix/interface/main.py
index ceb14c26..06966f4c 100644
--- a/strix/interface/main.py
+++ b/strix/interface/main.py
@@ -224,6 +224,7 @@ async def warm_up_llm(show_model_warning: bool = True) -> None:
                 request_timeout=llm.timeout,
                 prompt_cache=False,
                 extra_headers=settings.dedupe.extra_headers,
+                has_tools=False,
             )
             if deduper_extra:
                 merged = {**(deduper_settings.extra_args or {}), **deduper_extra}
diff --git a/strix/interface/scan_setup.py b/strix/interface/scan_setup.py
index 327769a3..1e795a5c 100644
--- a/strix/interface/scan_setup.py
+++ b/strix/interface/scan_setup.py
@@ -78,6 +78,7 @@ async def preflight_model_connection(
         request_timeout=resolved_settings.llm.timeout,
         prompt_cache=False,
         extra_headers=resolved_settings.llm.extra_headers,
+        has_tools=False,
     )
     await asyncio.wait_for(
         model.get_response(
diff --git a/strix/llm/compaction.py b/strix/llm/compaction.py
index 09d36454..e40caf6a 100644
--- a/strix/llm/compaction.py
+++ b/strix/llm/compaction.py
@@ -294,6 +294,7 @@ async def _summarize(model: str, prompt: str, max_tokens: int) -> str | None:
         request_timeout=llm.timeout,
         prompt_cache=False,
         extra_headers=llm.extra_headers,
+        has_tools=False,
     ).resolve(ModelSettings(max_tokens=max_tokens))
     try:
         response = (
diff --git a/strix/report/dedupe.py b/strix/report/dedupe.py
index f848a6d6..1cc0a66a 100644
--- a/strix/report/dedupe.py
+++ b/strix/report/dedupe.py
@@ -62,6 +62,7 @@ def _dedupe_model_settings(
         # must never receive the main endpoint's credentials. A dedicated model
         # gets its own DEDUPE_LLM_EXTRA_HEADERS instead.
         extra_headers=dedupe.extra_headers if dedupe.model else llm.extra_headers,
+        has_tools=False,
     )
     extra = _dedupe_extra_args(dedupe)
     if extra:
diff --git a/tests/test_inputs.py b/tests/test_inputs.py
index ed233262..7dea64be 100644
--- a/tests/test_inputs.py
+++ b/tests/test_inputs.py
@@ -299,6 +299,16 @@ def test_make_model_settings_forces_required_for_anyllm_routed_openai_model() ->
     assert settings.tool_choice == "required"
 
 
+def test_make_model_settings_disables_parallel_tool_calls_by_default() -> None:
+    assert make_model_settings("none", model_name="gpt-4o").parallel_tool_calls is False
+
+
+def test_make_model_settings_omits_parallel_tool_calls_without_tools() -> None:
+    settings = make_model_settings("none", model_name="gpt-4o", has_tools=False)
+
+    assert settings.parallel_tool_calls is None
+
+
 def test_make_model_settings_sets_request_timeout() -> None:
     settings = make_model_settings(
         "none",