← Back to Model Beat
Models·Sep 28·all news from September 28, 2026

Alibaba Qwen Releases Qwen-Audio-3.1-Realtime: A Full-Duplex Voice Model Trained to Think, Act, and Decide When to Speak

Alibaba has released Qwen-Audio-3.1-Realtime, a full-duplex voice model capable of reasoning, using external tools, and managing conversational timing. The update improves task success rates to 82% and significantly reduces unnecessary interruptions by background speech to 13%, marking a technical refinement in how the model handles live audio interactions. It is currently available for integration through the QwenCloud API.

Covered by 2 sources · 3 articles

Related stories

ModelsThe Lenfest Institute grows landmark program with expanded OpenAI supportSep 28 · 95 sourcesModelsGemini 4 Argon: our next era of frontier intelligenceSep 30 · 15 sourcesModelsDeepSeek Unveils Huawei AI Chip Tools That May Replace Nvidia’sSep 30 · 19 sourcesModelsIntroducing GPT-6.1 SolSep 29 · 6 sources