New open-source voice model listens nonstop and decides every 0.4 seconds whether to speak or stay silent

A new open-source voice model named Audio Interaction has been launched, capable of interacting without waiting for the end of a recording. This model translates, transcribes, chats, and captures everyday sounds like coughing in a continuous flow. Unlike GPT-4o and Qwen3.5-Omni, it decides every 0.4 seconds whether to speak or remain silent. The model's code, weights, and loading instructions are available on GitHub under the Apache 2.0 license, with a promise to provide training data later.