As reported by Alibaba, the Qwen team has just rolled out a speech recognition model named Qwen3-ASR-Flash. Based on the Qwen3-Omni framework, this model is capable of automatically detecting and transcribing speech in 11 languages, encompassing Chinese, English, Arabic, French, and more. Additionally, it is well-suited for handling various Chinese dialects as well as different regional accents of English.