From d8b385b7eac4724425f75bfe7a7ae3f477dd8d5d Mon Sep 17 00:00:00 2001 From: ffggs <314137448@qq.com> Date: Sun, 24 May 2026 00:03:19 +0800 Subject: [PATCH] [Bugfix][Frontend] Fix input_audio parsing when uuid is present (#43414) Signed-off-by: ffggs <314137448@qq.com> Co-authored-by: Cyrus Leung --- vllm/entrypoints/chat_utils.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/vllm/entrypoints/chat_utils.py b/vllm/entrypoints/chat_utils.py index e3b727597d..e4e44bc01a 100644 --- a/vllm/entrypoints/chat_utils.py +++ b/vllm/entrypoints/chat_utils.py @@ -1517,7 +1517,7 @@ def _parse_chat_message_content_mm_part( audio_url = audio_url.get("url", None) return "audio_url", audio_url if part.get("input_audio") is not None: - input_audio_params = cast(dict[str, str], part) + input_audio_params = _InputAudioParser(part).get("input_audio", None) return "input_audio", input_audio_params if "video_url" in part: video_params = cast(CustomChatCompletionContentSimpleVideoParam, part)