AI Radar
Top stories
Importance 2 of 5Covered by 1 source

Haru-4B released as open full-duplex spoken dialogue base model

Higurashi AI released Haru-4B, a multilingual full-duplex spoken dialogue foundation model that listens and speaks at the same time. It processes one audio frame every 133 ms and is released as a pretrained base model for research.

Why it matters

Full-duplex speech models point toward more natural voice agents for phone and customer service use.