# \#realtime **URL:** https://community.livekit.io/tag/realtime/28.md [Latest](https://community.livekit.io/latest.md) · [Categories](https://community.livekit.io/categories.md) · [Tags](https://community.livekit.io/tags.md) --- ## [Qwen3.8-Omni realtime on LiveKit: Inference, Agent Builder, or plugin?](https://community.livekit.io/t/qwen3-8-omni-realtime-on-livekit-inference-agent-builder-or-plugin/2316)
**Author:** [@Abbas\_Mohammed](https://community.livekit.io/u/Abbas_Mohammed)\ **Replies:** 2\ **Last updated:** [October 21, 2026, 10:18am UTC](https://community.livekit.io/t/qwen3-8-omni-realtime-on-livekit-inference-agent-builder-or-plugin/2316 "2026-10-21T10:18:05Z")
Hi! We’re building a phone-ordering voice agent for restaurants in Iraq (Iraqi Arabic). Gemini 3.8 Live through LiveKit Inference works for us, but we’d like to try Alibaba’s Qwen3.8-Omni realtime (speech-to-speech) beca… --- ## [Looking for a Realtime Voice AI Engineer / Consultant — LiveKit, Vapi, Retell + Stateful Backend](https://community.livekit.io/t/looking-for-a-realtime-voice-ai-engineer-consultant-livekit-vapi-retell-stateful-backend/2304)
**Author:** [@alex3](https://community.livekit.io/u/alex3)\ **Replies:** 4\ **Last updated:** [October 20, 2026, 5:24pm UTC](https://community.livekit.io/t/looking-for-a-realtime-voice-ai-engineer-consultant-livekit-vapi-retell-stateful-backend/2304 "2026-10-20T17:24:17Z")
I’m looking for an experienced realtime voice engineer for a paid architecture/code review of an existing AI receptionist. We originally built the voice runtime on LiveKit Agents. The business side of the system is alre… --- ## [Per-session CPU burst from a plugin's native model being built lazily on the first audio frame (AI\_Cousitcs) — any pattern for prewarmable per-session init?](https://community.livekit.io/t/per-session-cpu-burst-from-a-plugins-native-model-being-built-lazily-on-the-first-audio-frame-ai-cousitcs-any-pattern-for-prewarmable-per-session-init/2180)
**Author:** [@Paras\_Agarwal](https://community.livekit.io/u/Paras_Agarwal)\ **Replies:** 1\ **Last updated:** [October 9, 2026, 8:00am UTC](https://community.livekit.io/t/per-session-cpu-burst-from-a-plugins-native-model-being-built-lazily-on-the-first-audio-frame-ai-cousitcs-any-pattern-for-prewarmable-per-session-init/2180 "2026-10-09T08:00:45Z")
Description: A noise-cancellation plugin we use builds its native model lazily on each session’s first audio frame rather than ahead of time, causing a short but real CPU burst at the start of every session. Looking for… --- ## [Issue: RealtimeModelMetrics.ttft is always logged as 0.0 for gemini 3.8 live on vertex ai](https://community.livekit.io/t/issue-realtimemodelmetrics-ttft-is-always-logged-as-0-0-for-gemini-3-8-live-on-vertex-ai/2200)
**Author:** [@Harshita\_Sukumar\_Patil](https://community.livekit.io/u/Harshita_Sukumar_Patil)\ **Replies:** 2\ **Last updated:** [October 7, 2026, 2:28pm UTC](https://community.livekit.io/t/issue-realtimemodelmetrics-ttft-is-always-logged-as-0-0-for-gemini-3-8-live-on-vertex-ai/2200 "2026-10-07T14:28:27Z")
@livekit @LiveKit-Community @livekitteams @darryncampbell @CWilson @Muhammad\_Usman\_Bashir, @Akiniesta Hi LiveKit Team, We are observing that Time To First Token (ttft) is consistently logging as 0.0 ms when using the g… --- ## [Gemini-3.8-live rejected with vertexai=True in livekit-plugins-google, although Vertex AI serves it (EU multi-region)](https://community.livekit.io/t/gemini-3-8-live-rejected-with-vertexai-true-in-livekit-plugins-google-although-vertex-ai-serves-it-eu-multi-region/2179)
**Author:** [@Akiniesta](https://community.livekit.io/u/Akiniesta)\ **Replies:** 7\ **Last updated:** [October 6, 2026, 3:06pm UTC](https://community.livekit.io/t/gemini-3-8-live-rejected-with-vertexai-true-in-livekit-plugins-google-although-vertex-ai-serves-it-eu-multi-region/2179 "2026-10-06T15:06:32Z")
Hi LiveKit community, We run a realtime voice agent on Gemini Live (google.realtime.RealtimeModel), currently gemini-3.1-flash-live-preview through the Gemini Developer API. We want to move it to Vertex AI in the EU mul… --- ## [Was it us or them? Answering "the live froze for me" with evidence](https://community.livekit.io/t/was-it-us-or-them-answering-the-live-froze-for-me-with-evidence/2164)
**Author:** [@Arkadiusz\_Kubaczkowski](https://community.livekit.io/u/Arkadiusz_Kubaczkowski)\ **Replies:** 3\ **Last updated:** [October 1, 2026, 9:13am UTC](https://community.livekit.io/t/was-it-us-or-them-answering-the-live-froze-for-me-with-evidence/2164 "2026-10-01T09:13:24Z")
When a viewer says “the live froze for me”, someone has to answer a hard question: was it us, or them? Most live-video teams can’t answer it quickly. Server logs look clean because the servers did their job, dashboards … --- ## [Turn-taking on real SIP calls: what cut my median reply from 2.44 s to 1.81 s (plus a question on short answers)](https://community.livekit.io/t/turn-taking-on-real-sip-calls-what-cut-my-median-reply-from-2-44-s-to-1-81-s-plus-a-question-on-short-answers/2139)
**Author:** [@Mohammed\_Usmani](https://community.livekit.io/u/Mohammed_Usmani)\ **Replies:** 2\ **Last updated:** [September 29, 2026, 2:56pm UTC](https://community.livekit.io/t/turn-taking-on-real-sip-calls-what-cut-my-median-reply-from-2-44-s-to-1-81-s-plus-a-question-on-short-answers/2139 "2026-09-29T14:56:11Z")
I’ve been measuring a LiveKit Agents phone agent (Python, SIP, Deepgram, Gemini, Cartesia) turn by turn, and wanted to share a few findings in case they help others, plus one open question. What moved the numbers (agent… --- ## [Voice Live API connector](https://community.livekit.io/t/voice-live-api-connector/351)
**Author:** [@bruno.wlodarczyk](https://community.livekit.io/u/bruno.wlodarczyk)\ **Replies:** 9\ **Last updated:** [September 23, 2026, 7:21am UTC](https://community.livekit.io/t/voice-live-api-connector/351 "2026-09-23T07:21:12Z")
Hi all, are there plans to include this plugin from this PR in next releases? --- ## [Gemini Realtime API error: 1007 None. The audio content type (CONTENT\_TYPE\_AUDIO) is not supported for this model](https://community.livekit.io/t/gemini-realtime-api-error-1007-none-the-audio-content-type-content-type-audio-is-not-supported-for-this-model/1908)
**Author:** [@Peace](https://community.livekit.io/u/Peace)\ **Replies:** 6\ **Last updated:** [September 18, 2026, 10:13am UTC](https://community.livekit.io/t/gemini-realtime-api-error-1007-none-the-audio-content-type-content-type-audio-is-not-supported-for-this-model/1908 "2026-09-18T10:13:59Z")
1. Error Hi, I had reported this exact same error a while back (Kindly view here)and since then it was working temporarily 2. Context My observations remain same that I am still not able to figure out when and due to w… --- ## [OpenAI Realtime STT metrics recorded in Livekit](https://community.livekit.io/t/openai-realtime-stt-metrics-recorded-in-livekit/2066)
**Author:** [@Benjamin\_Lowe](https://community.livekit.io/u/Benjamin_Lowe)\ **Replies:** 2\ **Last updated:** [September 17, 2026, 1:56pm UTC](https://community.livekit.io/t/openai-realtime-stt-metrics-recorded-in-livekit/2066 "2026-09-17T13:56:47Z")
When using OpenAI realtime model with whisper STT, we are billed for STT seperately but livekit currently records no metadata for cost estimation, I made this PR to addresses this obervavility hole - would appreciate rev… --- ## [Self-hosted LiveKit for a university, which cloud/bare-metal provider fits this bandwidth-heavy load?](https://community.livekit.io/t/self-hosted-livekit-for-a-university-which-cloud-bare-metal-provider-fits-this-bandwidth-heavy-load/1998)
**Author:** [@Mohammad\_Khdour](https://community.livekit.io/u/Mohammad_Khdour)\ **Replies:** 3\ **Last updated:** [September 7, 2026, 5:18pm UTC](https://community.livekit.io/t/self-hosted-livekit-for-a-university-which-cloud-bare-metal-provider-fits-this-bandwidth-heavy-load/1998 "2026-09-07T17:18:50Z")
Hi everyone, planning a self-hosted LiveKit deployment for live university classes and would love input from anyone running LiveKit at a similar scale. Our setup: Use case: live classes + on-demand recording (Egress),… --- ## [Data residency of South African sensitive data -](https://community.livekit.io/t/data-residency-of-south-african-sensitive-data/1999)
**Author:** [@Lee](https://community.livekit.io/u/Lee)\ **Replies:** 4\ **Last updated:** [September 4, 2026, 12:59pm UTC](https://community.livekit.io/t/data-residency-of-south-african-sensitive-data/1999 "2026-09-04T12:59:10Z")
Hi LK team, This question is an adapted version of this ( Data residency for Deepgram STT when using LiveKit Inference in India region ) where i am handling financial data for a company and supplying a service to them b… --- ## [Agent shows persistent "ERROR" status — zero dispatch attempts logged, spanning multiple hours](https://community.livekit.io/t/agent-shows-persistent-error-status-zero-dispatch-attempts-logged-spanning-multiple-hours/1938)
**Author:** [@Andrew\_Goldberg](https://community.livekit.io/u/Andrew_Goldberg)\ **Replies:** 9\ **Last updated:** [August 31, 2026, 8:49am UTC](https://community.livekit.io/t/agent-shows-persistent-error-status-zero-dispatch-attempts-logged-spanning-multiple-hours/1938 "2026-08-31T08:49:01Z")
We’re seeing consistent silent dispatch failure — rooms are created and callers join, but the agent is never dispatched to them at all, despite the token correctly requesting it. The agent’s status page shows a persisten… --- ## [Inaccurate Transcript Despite Correct Realtime Understanding](https://community.livekit.io/t/inaccurate-transcript-despite-correct-realtime-understanding/1829)
**Author:** [@royibernthal](https://community.livekit.io/u/royibernthal)\ **Replies:** 23\ **Last updated:** [August 27, 2026, 3:48pm UTC](https://community.livekit.io/t/inaccurate-transcript-despite-correct-realtime-understanding/1829 "2026-08-27T15:48:14Z")
I’m using the following AgentSessionOptions for realtime speech in Hebrew: { llm: new openai.realtime.RealtimeModel({ model: 'gpt-realtime-2.1', voice: 'marin', inputAudioTranscription: { model: 'gpt… --- ## [Unity Android Voice Chat Issues (Playback Route, Bluetooth, Samsung Low Microphone Volume)](https://community.livekit.io/t/unity-android-voice-chat-issues-playback-route-bluetooth-samsung-low-microphone-volume/1765)
**Author:** [@Alireza\_Karimy](https://community.livekit.io/u/Alireza_Karimy)\ **Replies:** 17\ **Last updated:** [August 24, 2026, 8:17am UTC](https://community.livekit.io/t/unity-android-voice-chat-issues-playback-route-bluetooth-samsung-low-microphone-volume/1765 "2026-08-24T08:17:02Z")
Unity Android Voice Chat Issues (Playback Route, Bluetooth, Samsung Low Microphone Volume) Hi everyone, I’m integrating LiveKit Unity SDK into a multiplayer game for Android, and I’ve run into several audio routing issu… --- ## [Agent didn't followup after tool call](https://community.livekit.io/t/agent-didnt-followup-after-tool-call/1332)
**Author:** [@royibernthal](https://community.livekit.io/u/royibernthal)\ **Replies:** 9\ **Last updated:** [August 12, 2026, 6:23pm UTC](https://community.livekit.io/t/agent-didnt-followup-after-tool-call/1332 "2026-08-12T18:23:16Z")
In a certain session, the agent didn’t followup after calling a tool - said it’ll do something, invoked the tool, then didn’t followup. Any attempts to nudge it by saying something resulted in the same behavior happening… --- ## [PSA: the FrameProcessor resample pattern silently discards ~16% of caller audio](https://community.livekit.io/t/psa-the-frameprocessor-resample-pattern-silently-discards-16-of-caller-audio/1821)
**Author:** [@troy](https://community.livekit.io/u/troy)\ **Replies:** 4\ **Last updated:** [August 11, 2026, 4:05am UTC](https://community.livekit.io/t/psa-the-frameprocessor-resample-pattern-silently-discards-16-of-caller-audio/1821 "2026-08-11T04:05:12Z")
A noise-cancellation FrameProcessor that resamples to its model’s rate and back must not force its output length to match the input frame. rtc.AudioResampler.push() returns variable-size chunks, so trimming the surplus a… --- ## [How to size self-hosted LiveKit for 200 concurrent classrooms × 50 participants (10,000 total) — what hardware/config do I need?](https://community.livekit.io/t/how-to-size-self-hosted-livekit-for-200-concurrent-classrooms-x-50-participants-10-000-total-what-hardware-config-do-i-need/1779)
**Author:** [@Mohammad\_Khdour](https://community.livekit.io/u/Mohammad_Khdour)\ **Replies:** 0\ **Last updated:** [July 31, 2026, 11:26pm UTC](https://community.livekit.io/t/how-to-size-self-hosted-livekit-for-200-concurrent-classrooms-x-50-participants-10-000-total-what-hardware-config-do-i-need/1779 "2026-07-31T23:26:59Z")
Hi everyone, I’m running self-hosted LiveKit (Docker Compose today, planning a multi-node cluster) for a university live-classroom platform. I need to plan hardware and configuration for this specific target, and I’d li… --- ## [Seeking strategies to optimize connection reliability and audio quality for users in strict GCC network environments experiencing aggressive UDP throttling](https://community.livekit.io/t/seeking-strategies-to-optimize-connection-reliability-and-audio-quality-for-users-in-strict-gcc-network-environments-experiencing-aggressive-udp-throttling/1655)
**Author:** [@Faraan\_Ahmed\_Hashmi](https://community.livekit.io/u/Faraan_Ahmed_Hashmi)\ **Replies:** 18\ **Last updated:** [July 28, 2026, 4:35am UTC](https://community.livekit.io/t/seeking-strategies-to-optimize-connection-reliability-and-audio-quality-for-users-in-strict-gcc-network-environments-experiencing-aggressive-udp-throttling/1655 "2026-07-28T04:35:14Z")
Hello everyone, We are running a production real-time 1-on-1 tutoring web application built on LiveKit Cloud (React frontend + Node.js backend). While our global users experience flawless connectivity, we are facing sig… --- ## [Gemini Realtime API error: 1007 None. The audio content type (CONTENT\_TYPE\_AUDIO) is not supported for this model configuration](https://community.livekit.io/t/gemini-realtime-api-error-1007-none-the-audio-content-type-content-type-audio-is-not-supported-for-this-model-configuration/1737)
**Author:** [@Peace](https://community.livekit.io/u/Peace)\ **Replies:** 3\ **Last updated:** [July 27, 2026, 10:47pm UTC](https://community.livekit.io/t/gemini-realtime-api-error-1007-none-the-audio-content-type-content-type-audio-is-not-supported-for-this-model-configuration/1737 "2026-07-27T22:47:06Z")
1. Error Receiving 1007 error from Gemini Live API 2. System Windows 11 3. Context The only model I am using is → REALTIME\_MODEL = “gemini-2.5-flash-native-audio-preview-12-2025” Crash occurrence is complicated becaus… --- ## [Proposed enhancements for livekit.agents.llm.RealtimeModelFallbackAdapter](https://community.livekit.io/t/proposed-enhancements-for-livekit-agents-llm-realtimemodelfallbackadapter/1738)
**Author:** [@Terry\_So](https://community.livekit.io/u/Terry_So)\ **Replies:** 0\ **Last updated:** [July 27, 2026, 2:40am UTC](https://community.livekit.io/t/proposed-enhancements-for-livekit-agents-llm-realtimemodelfallbackadapter/1738 "2026-07-27T02:40:14Z")
Hi @Long Chen We have report a bug and proposed enhancements in the livekit agent repo, which are related to the livekit.agents.llm.RealtimeModelFallbackAdapter introduced in the latest released version v1.6.6 of liveki… --- ## [AWS Nova Sonic RealtimeModel (amazon.nova-2-sonic-v1:0): metrics emitted repeatedly and ttftalways0.0](https://community.livekit.io/t/aws-nova-sonic-realtimemodel-amazon-nova-2-sonic-v1-0-metrics-emitted-repeatedly-and-ttftalways0-0/1665)
**Author:** [@Shivakant\_Yadav](https://community.livekit.io/u/Shivakant_Yadav)\ **Replies:** 2\ **Last updated:** [July 21, 2026, 12:34am UTC](https://community.livekit.io/t/aws-nova-sonic-realtimemodel-amazon-nova-2-sonic-v1-0-metrics-emitted-repeatedly-and-ttftalways0-0/1665 "2026-07-21T00:34:32Z")
@livekit Hi Livekit team, When using the AWS Nova Sonic realtime model (amazon.nova-2-sonic-v1:0) via the livekit-plugins-aws realtime integration, I’m seeing two problems with the RealtimeModel metrics events: Metri… --- ## [I put LiveKit on an ESP32-S3 + ReSpeaker XVF3800 — a full-duplex smart speaker that listens while it talks](https://community.livekit.io/t/i-put-livekit-on-an-esp32-s3-respeaker-xvf3800-a-full-duplex-smart-speaker-that-listens-while-it-talks/1560)
**Author:** [@Ruben\_Garcia](https://community.livekit.io/u/Ruben_Garcia)\ **Replies:** 16\ **Last updated:** [July 9, 2026, 2:33pm UTC](https://community.livekit.io/t/i-put-livekit-on-an-esp32-s3-respeaker-xvf3800-a-full-duplex-smart-speaker-that-listens-while-it-talks/1560 "2026-07-09T14:33:54Z")
Hi all :waving\_hand: I’ve been building Sebastian, a real, standalone smart speaker you can put on your desk — a hands-free voice assistant that runs LiveKit’s WebRTC stack on a bare ESP32-S3 microcontroller (not a Pi, … --- ## [Bugs and proposed enhancements related to the generate\_reply timeout following up actions](https://community.livekit.io/t/bugs-and-proposed-enhancements-related-to-the-generate-reply-timeout-following-up-actions/1508)
**Author:** [@Terry\_So](https://community.livekit.io/u/Terry_So)\ **Replies:** 3\ **Last updated:** [June 29, 2026, 1:20pm UTC](https://community.livekit.io/t/bugs-and-proposed-enhancements-related-to-the-generate-reply-timeout-following-up-actions/1508 "2026-06-29T13:20:38Z")
Hi @Long Chen We have raised bugs and proposed enhancements in the livekit agent repo, which are related to the generate\_reply timeout following up actions: Please tackle all of them and keep us posted … --- ## [Payload type filtering](https://community.livekit.io/t/payload-type-filtering/1521)
**Author:** [@sbarber](https://community.livekit.io/u/sbarber)\ **Replies:** 2\ **Last updated:** [June 29, 2026, 9:36am UTC](https://community.livekit.io/t/payload-type-filtering/1521 "2026-06-29T09:36:49Z")
It would be ideal if we could embed json or protofuf directly into the RTP feed, thus taking advantage of existing FEC and QoS mechanisms in forwarding, however the Livekit SFU filters on WebRTC-compatible payload types. … --- ## [Generate reply timeout for gemini-live-2.5-flash-native-audio realtime model](https://community.livekit.io/t/generate-reply-timeout-for-gemini-live-2-5-flash-native-audio-realtime-model/1505)
**Author:** [@Harshita\_Sukumar\_Patil](https://community.livekit.io/u/Harshita_Sukumar_Patil)\ **Replies:** 2\ **Last updated:** [June 28, 2026, 1:38am UTC](https://community.livekit.io/t/generate-reply-timeout-for-gemini-live-2-5-flash-native-audio-realtime-model/1505 "2026-06-28T01:38:45Z")
@livekit @LiveKit-Community @livekitteams @darryncampbell @CWilson @Muhammad\_Usman\_Bashir Hi LiveKit Team, We are experiencing intermittent timeouts while using the Gemini Live 2.5 Flash Native Audio realtime model via… --- ## [Livekit Gemini-Live Pause issue](https://community.livekit.io/t/livekit-gemini-live-pause-issue/1228)
**Author:** [@naahuja](https://community.livekit.io/u/naahuja)\ **Replies:** 9\ **Last updated:** [June 19, 2026, 5:01pm UTC](https://community.livekit.io/t/livekit-gemini-live-pause-issue/1228 "2026-06-19T17:01:54Z")
So we are using livekit + gemini live(2.5 native audio) , and 4/10 times we are facing a pausing issue , and we are getting this in logs : {“message”: “\_SegmentSynchronizerImpl.playback\_finished called before text/audio… --- ## [Audio Intelligence for Voice Agents](https://community.livekit.io/t/audio-intelligence-for-voice-agents/1457)
**Author:** [@Pawel\_Lach](https://community.livekit.io/u/Pawel_Lach)\ **Replies:** 1\ **Last updated:** [June 19, 2026, 7:44am UTC](https://community.livekit.io/t/audio-intelligence-for-voice-agents/1457 "2026-06-19T07:44:15Z")
While most STT engines can only transcribe speech to text, they are unable to hear or recognize anything else. That’s why we developed Tyto: to bring audio intelligence to Voice AI and give your agent context about the a… --- ## [LiveKit Cloud GStreamer Pipeline Crashing with Connection Timeout (~4 min mark)](https://community.livekit.io/t/livekit-cloud-gstreamer-pipeline-crashing-with-connection-timeout-4-min-mark/533)
**Author:** [@Tom\_Fanella](https://community.livekit.io/u/Tom_Fanella)\ **Replies:** 3\ **Last updated:** [June 18, 2026, 7:59pm UTC](https://community.livekit.io/t/livekit-cloud-gstreamer-pipeline-crashing-with-connection-timeout-4-min-mark/533 "2026-06-18T19:59:38Z")
Urgent: Intermittent CONNECTION\_TIMEOUT / action=Resume causing livekitwebrtcsink publisher pipeline crashes Hi LiveKit community, I wanted to flag a critically urgent issue we have been experiencing with our LiveKit i… --- ## [Session.say() hangs indefinitely (~20s+) when using Gemini RealtimeModel(gemini 3.1) as llm together with google.beta.GeminiTTS as tts; session.input.set\_audio\_enabled() / aec\_warmup\_duration do not reliably block user interruption in this configuration](https://community.livekit.io/t/session-say-hangs-indefinitely-20s-when-using-gemini-realtimemodel-gemini-3-1-as-llm-together-with-google-beta-geminitts-as-tts-session-input-set-audio-enabled-aec-warmup-duration-do-not-reliably-block-user-interruption-in-this-configuration/1448)
**Author:** [@Jayesh\_Shinde](https://community.livekit.io/u/Jayesh_Shinde)\ **Replies:** 3\ **Last updated:** [June 18, 2026, 10:03am UTC](https://community.livekit.io/t/session-say-hangs-indefinitely-20s-when-using-gemini-realtimemodel-gemini-3-1-as-llm-together-with-google-beta-geminitts-as-tts-session-input-set-audio-enabled-aec-warmup-duration-do-not-reliably-block-user-interruption-in-this-configuration/1448 "2026-06-18T10:03:41Z")
Summary When using llm=google.beta.realtime.RealtimeModel (gemini-3.1-flash-live-preview) together with a separate tts=google.beta.GeminiTTS(gemini-3.1-flash-tts-preview) instance on the same AgentSession, a sequence of … [Next page](https://community.livekit.io/tag/realtime/28.md?match_all_tags=true&page=1&tags%5B%5D=realtime)