Qwen Alibaba Cloud: Qwen3 ASR Flash

Qwen multilingual speech recognition model for pre-recorded audio. Transcribes across 27 documented language codes with automatic language detection, free-text context biasing for domain terms and hotwords, optional inverse text normalization, and an always-on emotion label returned alongside the transcript.