AI BriefingQwen Audio Agent Released
Key point
Qwen has released a voice agent runtime that maintains conversations during background tasks.
Details
Qwen Audio Agent is a real-time voice runtime designed to separate voice conversations from background agent tasks, allowing users to check progress or cancel tasks while searching, coding, or invoking tools.
- Provides WebUI, terminal TUI, and desktop voice orb interfaces.
- Voice Gateway and TaskManager manage task lifecycles and states.
- Integrates with ACP-based backend agents such as OpenCode, Qwen Code, Kimi Code, Codex, and Claude Code.
- Reuses model, tool, MCP, skill, and authentication configurations from installed backends, enabling asynchronous execution of multiple tasks.
- The latest version is v1.8.3, requiring Node.js 22.22.2 or higher and npm 10 or higher.
The default voice frontend transmits audio and conversations via the DashScope real-time voice service. To avoid using cloud APIs, it can be replaced with a Hugging Face-based local speech-to-speech frontend.
Granting full permissions to the backend enables command execution and file modification, so it should only be used in trusted projects. It is also necessary to avoid exposing the local gateway to external networks.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Whiteboard Launches Open-Source VS Code Fork for Human-AI Collaborative Software Design
OpenComputer Hosts 'Ship Video' Agent Using Claude Opus 5.5 to Generate Explainer Videos from Code
Qwen Intelligence Launches Three Mobile AI Agents and Open Benchmark Suite