Qwen-Audio-3.1-ASR with Improved Speaker Diarization
October 19, 2026
Qwen-Audio-3.1-ASR expands audio-language capabilities to include structured speaker-diarized transcription with timestamps and overlapping speech handling. It supports 30 languages and 16 Chinese dialects, plus instruction-controlled transcription polishing.
HOW THIS AFFECTS YOU
●
builderYou can integrate more robust, multi-speaker audio understanding and transcription into your products with structured speaker labels.