Ali Thousand Questions Open Source Speech Recognition Model Qwen3-ASR

Ali Qwen announced the official open source of the Qwen3-ASR series models. According to the introduction, Qwen3-ASR is a series of powerful speech recognition models developed by Qwen, including two powerful and comprehensive speech recognition models Qwen3-ASR-1.7B and Qwen3-ASR-0.6B, as well as an innovative speech forced alignment model Qwen3-ForcedAligner-0.6B. The speech recognition models in the Qwen3-ASR series support language recognition and speech recognition for 52 languages and dialects.