Today, Alibaba Cloud Qoder launched QoderVoice, a real-time voice interaction agent powered by the full-duplex voice model Qwen-Audio-3.0-Realtime. Users can wake up the agent by voice at any time, issue commands, and discuss solutions. The agent responds in real time and executes tasks.
The official detailed introduction is as follows:
Voice commands and task execution proceed simultaneously
Users can click the voice entry on the main interface in Qoder Quest mode to summon the floating voice assistant window from any interface. After issuing a voice command, Qoder automatically creates a task, calls tools, and executes it in the background. Users do not need to remain on the page waiting for the result; they can interrupt, ask follow-up questions, or adjust their requirements at any time during the conversation.
Supports conversational interaction, enabling multi-turn negotiation for complex tasks
When handling complex tasks, Qoder first provides an execution plan. Users can supplement their requirements or adjust the direction by voice, and Qoder optimizes and executes the plan based on the feedback. For example, when a user says, “Help me refactor this module,” Qoder provides multiple options. The user can then say, “Option B is better, but add caching,” and Qoder will immediately make the adjustment and continue executing.
Now available on the Qoder international version, with a 3-day free trial
Qoder Voice is now available on the Qoder international version. It is initially open to individual subscription, individual trial, and enterprise subscription users, with a 3-day limited-time free trial. The Qoder CN version will be launched soon.