Qwen-Audio-3.1-ASR
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 1
- Qwen-Audio-3.1-ASR technical report claims improved diarized transcription
A 31-page arXiv technical report submitted on 20 October 2026 presents Qwen-Audio-3.1-ASR, a family of audio-language models the authors describe as building on Qwen-Audio-3.0-ASR with improved speech recognition and expanded speaker-diarized transcription and general audio understanding. The authors say the family retains support for 30 languages and 16 Chinese dialects, adds instruction-controlled features such as historical-context conditioning, hierarchical hotword customization and native transcription polishing, and includes variants for asynchronous file transcription, low-latency streaming, interactive voice input and non-streaming recognition. They also claim speaker-diarized output with speaker labels, timestamps and text that stay consistent across turns and represent overlapping speech, plus captioning, temporal grounding, sound event detection, audio question answering and multiple-choice reasoning. Evaluations are said to cover conventional ASR, multilingual and dialectal speech, diarized transcription and audio understanding, but the listing provides no numeric results or independent measurements.
Hugging Face · Papers · 🔥 0
Experience and discussion from the community
Share my Qwen-Audio-3.1-ASR experienceAsk about Qwen-Audio-3.1-ASR
Nobody has shared their experience with Qwen-Audio-3.1-ASR yet.