Turn detector: cloud with automatic local fallback

Patch Changes

• feat(core): audio end-of-turn detection with cloud → local fallback (AGT-2520) - #1719 (@chenghao-mou)

  • New inference.TurnDetector: WebSocket cloud EOT transport (version: 'v1', model name turn-detector-v1) with automatic fallback to the local native model (version: 'v1-mini', model name turn-detector-v1-mini) via @livekit/local-inference. Auto-selects 'v1' when LIVEKIT_REMOTE_EOT_URL is set, 'v1-mini' otherwise. The version is the constructor knob; telemetry/billing report the full model name via detector.model.
  • The local EOT model runs in the shared inference process (the same InferenceProcExecutor the text turn detector uses), loaded once per worker host (~138 MB) instead of in every job worker. The runner is registered by default when the native binding is available, so the inference process spawns on worker startup; on platforms where the binding can't load, local EOT degrades to a positive-default prediction and the worker still starts. (This is a JS-specific divergence from Python, which keeps EOT in-process and relies on forkserver COW sharing.)
  • No prewarm helpers: EOT auto-warms in the inference process; the in-process silero VAD lazy-loads on first stream. (The inference.prewarm* helpers added during development were removed before release.)
  • New inference.VAD (local-only streaming VAD via @livekit/local-inference).
  • AgentSession now auto-provisions a bundled silero VAD when vad is omitted (isDefault=true). Pass vad: null to opt out.
  • livekit-plugins-silero is deprecated; pass vad: null to opt out of the bundled default, or use inference.VAD({ model: 'silero', ... }) to customise.
  • livekit-plugins-livekit turn detector is deprecated in favor of inference.TurnDetector.
  • Endpointing defaults are now detector-aware: when the resolved turn detector is a streaming ("audio model") detector — the bundled default — unset endpointing keys fall back to tighter defaults (minDelay: 300, maxDelay: 2500) instead of the legacy 500/3000. Non-streaming modes (vad/stt/manual/realtime_llm, or turnDetection: null) keep the legacy defaults. Explicit user keys are tracked as sparse overrides and re-resolved per agent activity, so different agents in one session can use different detectors and runtime updateOptions changes survive handoffs.
  • New EOTInferenceMetrics and EOTModelUsage; new telemetry span attributes (lk.eou.source, lk.eou.from_cache, lk.eou.detection_delay); new eot_prediction event forwarded over remote sessions.
  • Requires @livekit/protocol >= 1.46.5 (exposes the AgentInference message namespace used by the cloud transport, including the server-provided SessionCreated default thresholds).

• Updated dependencies [27a6e829350c13fcdca533d68f864bebda70de89, 9cc7215bc08c34f24b5d9f7f8fbe754d7e67c267, ed2364ad105d7fde9baccc463a7bdbffa6a1699c, ed2364ad105d7fde9baccc463a7bdbffa6a1699c, 27a6e829350c13fcdca533d68f864bebda70de89, e64698c2e67048ff577d5024488929193d0b60e4, ec4a2a48d7ba1f6c20a86303b264188fa47fae0d, e1acca813568869fd345b5eee16be211e8595d9b, bb8e6251354062714e39ae5a44244e1ef65b385b, ed2364ad105d7fde9baccc463a7bdbffa6a1699c]:

  • @livekit/agents@1.4.7