字节发布 SeedRealtime 音视频全双工大模型，实现多模态自然交互

AI 摘要

字节跳动旗下 Seed Research 正式发布 SeedRealtime，一款原生音视频全双工大语言模型。该模型采用统一架构原生融合音频、视频和文本，实现实时多模态交互，具备联合音视频理解、主动交互和自然对话节奏三大突破。端到端评估显示，相比级联模型，其音视频对话节奏问题减半，并已在行业率先实现大规模部署。

正文节选

Today, we are officially launching SeedRealtime, a native audio-visual full-duplex LLM. As a key step toward omni-modal interaction, SeedRealtime uses a unified architecture to natively fuse audio, video, and text, enabling real-time interaction over continuous multimodal streams and delivering a brand-new "watch, listen, and speak" experience. It achieves three core breakthroughs: Joint audio-visual understanding: Native support for the deep fusion of audio, visual, and temporal information. Th

发布时间：—

抓取时间：2026-08-17 03:43

来源机构：ByteDance Seed