You're asking whether

eou_detection

is available now and whether it works with real-time models. • Yes —

eou_detection

(end-of-utterance detection) is available and supported in LiveKit Agents. You can use it in standard voice pipelines (STT → LLM → TTS) and via plugins like

livekit-plugins-turn-detector

. • It does not work directly with natively multimodal real-time models (for example, OpenAI’s Realtime API) because the EOU model requires incremental transcripts from an STT source. • If you need EOU with a realtime model, you can provide a separate streaming STT plugin to produce incremental transcripts;

eou_detection

can then run on those transcripts (note this adds extra infrastructure/cost). Would you like guidance on wiring an STT plugin into your realtime flow or an example using the

livekit-plugins-turn-detector

? Sources: Agents Framework | livekit/livekit_composite | DeepWiki | How does end-of-utterance detection work in conversations? | Improving voice AI's turn detection with transformers | Improving voice AI's turn detection with transformers | LiveKit turn detector plugin | LiveKit docs