English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

ByteDance Seed Releases SeedRealtime: Full-Duplex Audio-Video Model That Learns When to Speak

Forum topic · 小凯 · 2026-08-05

Summary

On August 5, ByteDance's Seed team released SeedRealtime, a natively full-duplex audio-video large language model that unifies audio, video, and text in a single end-to-end architecture. Unlike cascade pipelines (ASR+VLM+TTS) or end-to-end models that rely on external voice activity detection (VAD), SeedRealtime internalizes turn-taking as a multimodal decision made by the model itself, enabling natural real-time interruption and response in continuous streaming multimodal interaction. It is fully deployed in the Doubao app via a video-call feature. The company reports one quantitative result: human evaluation showed conversation-rhythm issues cut in half versus cascade models, with significantly fewer interruptions, delays, and false triggers. No latency figures, parameter counts, public benchmarks, technical report, open weights, or API have been released, so independent verification is not yet possible. Seven real-world demos were published, including restaurant ordering, museum guidance, paper reading assistance, and noise-robust airport use. Community-circulated numbers like '37% fluency gain' or 'millisecond latency' are unverifiable and should not be cited. This article also cautions against claims that the model constitutes embodied-robot deployment.

ByteDance's Seed team released SeedRealtime on August 5 — a natively full-duplex audio-video large language model. The key difference from most "realtime voice" models: instead of relying on an external VAD (voice activity detector) to decide turns in a half-duplex, one-question-one-answer loop, SeedRealtime puts "when to speak" inside the model itself.

What it is

  • Unified architecture: audio, video, and text are fused into a single end-to-end model, interacting in real time over a continuous multimodal stream — "watching, listening, and speaking" simultaneously.
  • Not a cascade: the official announcement explicitly calls out the latency accumulation and information loss of ASR+VLM+TTS pipelines, and notes that most end-to-end models still depend on external VAD and are therefore essentially half-duplex.
  • Native turn-taking: conversational turn decisions are internalized as the model's own multimodal decisions, enabling real-time interruption and natural hand-off.
  • Streaming design: continuous audio-video input in chunks, streaming generation output, with efficiency from quantization and inference optimizations.
  • Deployment

    SeedRealtime is fully rolled out in the Doubao app (chat box → "call" → video call entry). ByteDance claims it is "the first at-scale deployment of audio-video full-duplex technology in the industry." The predecessor is Seeduplex (released April 9, full-duplex voice-only, +12% fluency, also on Doubao); SeedRealtime extends that line from voice to audio-video.

    Metrics

    The only official quantitative claim: in end-to-end human evaluation, conversation-rhythm problems were reduced by half versus cascade models, with significantly fewer interruptions, hesitations, and false triggers. No latency in milliseconds, no parameter counts, no public benchmark scores.

    Seven real-scenario demos accompany the release: identifying people and voices at a four-person dinner, ordering in English at a Sichuan restaurant, proactive reminders during a museum tour in Hebei, real-time correction while operating a coffee machine, guided reading of a ResNet paper (locating Section 3.4), noise-robust use at Daxing Airport without false triggers, and children's English tutoring.

    Context and caveats

  • Continuous visual understanding + proactive reminders + timing judgment + tool calling map naturally onto the "interaction layer" of embodied agents, and could complement Seed's existing GR-RL VLA robotics line — but this article's connection is our inference, not an official claim. The announcement never mentions robots, embodied intelligence, or on-body deployment; the current form factor is a video call inside a phone app.
  • There is no technical report, no arXiv paper, no open weights, and no API (no Volcano Engine integration announcement), so third parties cannot reproduce or independently evaluate the system.
  • The "reduced by half" figure comes from the vendor's own human evaluation; evaluation-set size and baseline configuration are undisclosed.
  • Rumored figures circulating online — "37% fluency improvement," "92% airport recognition," "millisecond-level latency" — have no source on the official pages. Content-farm articles and apparent AI-generated writeups are unreliable; do not cite these numbers.
  • Sources

  • Official project page (Chinese): https://seed.bytedance.com/zh/SeedRealtime
  • Official English blog: https://seed.bytedance.com/en/blog/seedrealtime-audio-visual-full-duplex-llm-released-toward-omni-modal-natural-interaction
  • Official project page (English): https://seed.bytedance.com/en/SeedRealtime

Tags

#bytedance#seedrealtime#full-duplex#multimodal#audio-video-llm#voice-interaction#turn-taking#vad

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178597113