A high-throughput and memory-efficient inference and serving engine for LLMs
Fix parallel_tool_calls: null treated as false instead of default true (#44955)
Signed-off-by: factnn <166481866+factnn@users.noreply.github.com>
Zang Peiyu committed
Jun 16, 2026 at 05:14 UTC
25ee659db01f42747e87e784c139c0686f2cada6
1 file changed
+2
-2
vllm/entrypoints/serve/utils/tool_calls_utils.py
+2
-2
@@ -19,9 +19,9 @@ _ChatCompletionResponseChoiceT = TypeVar(
19
def maybe_filter_parallel_tool_calls(
20
choice: _ChatCompletionResponseChoiceT, request: ChatCompletionRequest
21
) -> _ChatCompletionResponseChoiceT:
22
- """Filter to first tool call only when parallel_tool_calls is False."""
22
+ """Filter to first tool call only when parallel_tool_calls is explicitly False."""
23
24
- if request.parallel_tool_calls:
24
+ if request.parallel_tool_calls is not False:
25
return choice
26
27
if isinstance(choice, ChatCompletionResponseChoice) and choice.message.tool_calls: