Alibaba Releases Qwen-Audio-3.0-ASR: 95% Accuracy in Medical Terms
智东西 · wechat · 2026-07-31
Alibaba's Qwen team released Qwen-Audio-3.0-ASR-Flash, a new speech recognition model. The update focuses on context consistency, industry-specific vocabulary, and custom hotwords. It achieves a 95.36% hit rate for medical terminology and over 90% accuracy in industrial and IT programming contexts.
Beyond raw accuracy, the model features built-in speech refinement capabilities. It automatically removes filler words, stutters, and self-corrections to output structured, written-quality text directly, matching the performance of traditional two-step pipelines. Additionally, it supports mixed recognition across 30 languages with an average semantic error rate of 17.09%. The model is now available via Alibaba Cloud's Bailian platform.
Related event: Alibaba Releases Qwen-Audio-3.0-ASR-Flash Model(3 posts)→
More from Models
- Google restricts Gemini Pro access for Google Plus subscribers, matching free tier — EmergencyLess8861 · 2026-10-16
- Meta releases Muse Spark Open, trained with vLLM-powered RL inference from day one — colinmcnamara · 2026-10-16
- Meta open-sources Muse Spark: runs on two Mac Studios, license bans passing output off as human-written — mawcopolow · 2026-10-16
- Golden Gate Claude explainer resurfaces: interpretability premise still holds — HarperSCarroll · 2026-10-16
- Hugging Face livestreams release of a new open-weight model — huggingface · 2026-10-16
- Step 5 Preview hits #2 on OpenRouter weekly leaderboard with 19.5T tokens in one week — StepFun_ai · 2026-10-16