ReleasesSep 29, 2026 · 11:581 source

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

From marktechpost.com

AI;DR

Alibaba's Qwen team released Qwen-Audio-3.1-Realtime, a full-duplex voice model that can reason, call tools, and decide when to speak, achieving an 82.0% task success rate on a τ-Voice adaptation and reducing replies to background speech from 73% to 13%. The model is available now as an API on QwenCloud.

Read the sourceSign in to suggest

Sources

#agentMarkdown

One ranked edition, every morning

The stories that matter today. Free, cancel anytime.

Telegram @aidr_today