ByteDance launches SeedRealtime, a large-scale audio and video full-duplex model.

ByteDance's Seed platform has launched SeedRealtime, a native full-duplex audio and video model. Based on a unified architecture, it natively integrates audio, video, and text to achieve real-time multimodal interaction of "watching, listening, and speaking simultaneously." The model possesses three core capabilities: joint audio and video understanding, proactive interaction, and smooth dialogue rhythm. End-to-end evaluation shows that dialogue rhythm issues are reduced by half compared to cascaded models.