
핵심 정리
- 1
ByteDance launched SeedRealtime, a full-duplex audio-video model.
- 2
The model processes audio, video, and text streams concurrently.
- 3
It facilitates real-time interactions, allowing simultaneous listening and responding.
- 4
SeedRealtime has been implemented in the Doubao app.
- 5
This launch marks the transition from a research demo to a consumer product.
- 6
It complements existing text and image generation capabilities at ByteDance.
관련 태그


