Hey all, I saw that Vision OS support was recently added to the client-sdk-swift, it looks like it uses the AR Kit CameraFrameProvider API to access the front facing camera. Does that mean we should request the com.apple.developer.arkit.main-camera-access.allow entitlement from apple to use this feature?

folks, this used to work in livekit-agents v0.7.2, post upgrade to v0.8.3, it's always None or "", im not sure which. how do i get the room_name now?

room_name = ctx.room.name
print("\n\nroom name: ", room_name)

Hey, I would like to make a an Android Local video chat app. My question: Is it possible that a client and server are one device and everyone in the network connects to this device and if the connection is lost, everyone should automatically connect to another device, which is then the server, so to speak.

On the server webhooks, it seems like whether or not the request is retried depends on the status code - ie. it seems to retry if we return a 500 code, but not a 404. Are there any details on which status codes will result in a retry?

Hello, I'd like to ask about the Ingress feature of Livekit. Does it support input with HTTPS streaming, similar to Icecast2 or Shoutcast?"

Context: We have an EKS cluster running with small

c6a.large

instances (2 vCPUs, 4 GiB RAM). The cluster is configured to scale up to a maximum of three instances when CPU utilization exceeds 60%. Load Testing Setup:

lk load-test --room loadTest01 --duration 30m --subscribers 1500 --video-resolution low --num-per-second 1000
lk load-test --room loadTest01 --duration 20m --video-publishers 1 video-resolution low --no-simulcast

We replicated this test configuration across two additional rooms (

loadTest02

and

loadTest03

), each with the same load of 1,500 subscribers and 1 publisher, running concurrently on our EKS cluster. The attached screenshot illustrates the CPU utilization graph for each instance in the cluster during the tests. Questions: 1. Room Allocation to Servers: Given that our LiveKit EKS cluster is configured with Redis (presumably in a distributed setup) and understanding that, by design, a room can only reside on a single server, how does LiveKit determine which server instance will host a particular room? 2. Manual Pod Allocation: In a scenario where we need to handle 9,000 users (distributing 3,000 users per instance), is it possible to specify which pod will host a particular room? If so, how can this be configured? 🙏

Hi. Does LiveKit support setting codecs on room level beside global room configuration ? I need for some rooms allow opus/vp9 but for others opus/vp9/vp8 can be used

In the latest version of

livekit_client: 2.2.3

and

flutter_webrtc: ^0.11.6

it is not possible to switch audio device on Android using

room.setAudioOutputDevice(device)

. It seems like its the OS who decide what audio device to use since I can't switch from either speakerphone (if no bluetooth headset is connected) or bluetooth headset.

Hey i am getting this error while running the agent server in local

(venv) yash@Yashs-MacBook-Air livekit % python3 main.py start                           
{"message": "starting worker", "version": "0.8.3", "timestamp": "2024-08-06T10:38:13.227942+00:00"}
{"message": "failed to connect to livekit, retrying in 0s: Cannot connect to host esa-irceuaug.livekit.cloud:443 ssl:True [SSLCertVerificationError: (1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1000)')]", "timestamp": "2024-08-06T10:38:13.410985+00:00"}
{"message": "failed to connect to livekit, retrying in 2s: Cannot connect to host esa-irceuaug.livekit.cloud:443 ssl:True [SSLCertVerificationError: (1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1000)')]", "timestamp": "2024-08-06T10:38:13.459755+00:00"}
{"message": "failed to connect to livekit, retrying in 4s: Cannot connect to host esa-irceuaug.livekit.cloud:443 ssl:True [SSLCertVerificationError: (1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1000)')]", "timestamp": "2024-08-06T10:38:15.520275+00:00"}

After properly providing the enviroment variables its still asking for ssl certificates Can anyone help here ??

Hello:) after using JS in livekit, i now want to use flutter, i have this widget below that connects to Livekit because i see the device connected in the admin console eating data, i just cannot find how to connect the track to the VideoTrackRenderer, my code is below and the stuff commented out is what i have been trying, I would love it if someone could help out:)

Sorry, here is the code Future<void> _connectToRoom() async { try { _room = Room(); await _room!.connect(url, token); // _listener = _room!.createListener(); //await _room!.participant.listen(_onTrackSubscribed); // _room!.onParticipantConnected.listen(_onParticipantConnected); // _room!.onParticipantDisconnected.listen(_onParticipantDisconnected); } catch (error) { print('Failed to connect to room: $error'); } } // void _onParticipantConnected(ParticipantConnectedEvent event) { // event.participant.onTrackSubscribed.listen(_onTrackSubscribed); // event.participant.onTrackUnsubscribed.listen(_onTrackUnsubscribed); // } // void _onParticipantDisconnected(ParticipantDisconnectedEvent event) { // // Handle participant disconnect if needed // } void _onTrackSubscribed(TrackSubscribedEvent event) { if (event.track is VideoTrack) { setState(() { _videoTracks.add(event.track as VideoTrack); }); } } void _onTrackUnsubscribed(TrackUnsubscribedEvent event) { if (event.track is VideoTrack) { setState(() { _videoTracks.remove(event.track as VideoTrack); }); } } @override Widget build(BuildContext context) { return Container( width: widget.width, height: widget.height, child: _videoTracks.isNotEmpty ? VideoTrackRenderer(_videoTracks.first) : Center(child: CircularProgressIndicator()), ); } @override void dispose() { _room?.disconnect(); super.dispose(); } }

Hi team! I set up the new VoiceAssistant (amazing work) on a python backend similar to the example you published on GitHub. I am now trying to render in Next.js the chat message history similar to how it is rendered on the LiveKit cloud playground (i.e. streamed transcription of user message & voice assistant reply). I tried wrapping the

<Chat/>

component within

<LiveKitRoom/>

after wrapping it in`<LayoutContextProvider/>`(throws LayoutContext error otherwise), as well as through useChat():

const { chatMessages: messages, send } = useChat();
  console.log(messages)

However, in both cases I can't get any messages returned although the voice assistant answers everything nicely on audio. Can you help point me in the right direction?

When we run

client.deleteRoom(videoRoomUUID).execute();

from our backend and mobile participants are currently connected to the room. Webhooks

room_finished

AND THEN

room_started

are trigger leaving us with a new and empty room every single time. When running web to web, we have no issue at all. We are using Flutter and

livekit-server 1.7.0

server-sdk-kotlin 0.5.9

Hi Team! I am wondering if there is a way to update tokens. If so where can I learn about it? For context: I am making a livekit demo where admin can create room and invite people and give them certain permissions. The generated token will then be: localhost:3000/room?permissionToken=....... The permission token is a token with only the permsions and room name. When a user goes to the link it generates a new token based on the permission token and adds identity of the user. If the admin kicks the user out nothing stops him from rejoining. I want to make the link with the permission token invalid. Is there a way to invalidate a token? Thank in advance for your time!

Hello I am trying to run multiple agents but getting following errors that address already in use- Do i need to run this on a different port or Am i missing some configuration? How can I run multiple agents? INFOlivekit.agentsstarting worker {"asctime": "2024-08-06 111320,817", "level": "INFO", "name": "livekit.agents", "message": "starting worker", "version": "0.7.1"} ERRORlivekit.agentsworker failed Traceback (most recent call last): File "/Users/projects/facedetection/lib/python3.10/site-packages/livekit/agents/cli/cli.py", line 227, in _worker_run await worker.run() File "/Users/projects/facedetection/lib/python3.10/site-packages/livekit/agents/worker.py", line 144, in run await asyncio.gather(self._worker_task(), self._http_server.run()) File "/Users/projects/facedetection/lib/python3.10/site-packages/livekit/agents/http_server.py", line 27, in run await site.start() File "/Users/projects/facedetection/lib/python3.10/site-packages/aiohttp/web_runner.py", line 119, in start self._server = await loop.create_server( File "/Users/anaconda3/lib/python3.10/asyncio/base_events.py", line 1519, in create_server raise OSError(err.errno, 'error while attempting ' OSError: [Errno 48] error while attempting to bind on address ('127.0.0.1', 8081): address already in use {"asctime": "2024-08-06 111320,821", "level": "ERROR", "name": "livekit.agents", "message": "worker failed", "exc_info": "Traceback (most recent call last):\n File \"/Users/projects/facedetection/lib/python3.10/site-packages/livekit/agents/cli/cli.py\", line 227, in _worker_run\n await worker.run()\n File \"Users/projects/facedetection/lib/python3.10/site-packages/livekit/agents/worker.py\", line 144, in run\n await asyncio.gather(self._worker_task(), self._http_server.run())\n File \"/Users/projects/facedetection/lib/python3.10/site-packages/livekit/agents/http_server.py\", line 27, in run\n await site.start()\n File \"/Users/projects/facedetection/lib/python3.10/site-packages/aiohttp/web_runner.py\", line 119, in start\n self._server = await loop.create_server(\n File \"/Users/anaconda3/lib/python3.10/asyncio/base_events.py\", line 1519, in create_server\n raise OSError(err.errno, 'error while attempting '\nOSError: [Errno 48] error while attempting to bind on address ('127.0.0.1', 8081): address already in use"}

I have tested the demo code that was present in docs, but i dont understand few things: 1. What really is LiveKit doing? Is it like a framework for Audio-Video, same as langchain is for LLMs? 2. I have noticed that it uses deepgram, is it really required? Doesn't LiveKit has it's own TTS and STT service? Isn't that what pricing is for? 3. I don't want to use rooms and all. Only api call and it should answer it. Is it possible? I find LiveKit interesting but i am unable to understand the pricing of it and use of deepgram and all.

Hello Team, facing same issue as @agreeable-caravan-96409 , i am trying to identify which user sent the chat message, but nothing shows up in the console window, please help, i added this @ctx.room.on("participant_connected") def on_participant_connected(participant: rtc.RemoteParticipant): logging.info( "participant connected: %s %s", participant.sid, participant.identity)

What instance size is recommended / are people using in production to run the VoiceAssistant agent? We're only handling about 10 concurrent users at a time.

Does livekit supports virtual background while streaming video?

Hello, Can anyone shed somelight on this variable "capacity" ? or please point to the technical documentation with some explanation. remote_video_stream = rtc.VideoStream(remote_track, format=rtc.VideoBufferType.RGB24, capacity=1)

Does anyone have an example iOS chat app that uses LiveKit, OpenAI, ElevenLabs and Firebase?

can anyone tell me what the the external egress custom recording template so i can use it from local server ?

template_base: // here use cloud template

Hello, team! I've managed to integrate livekit sip with asterisk. It works I can connect the voip client (like Zoiper) to livekit room, but when I try mute/unmute on voip client seems the livekit sip participant doesn't catch this state changes. Is it bug? Or livekit sip doesn't support mute/unmute sip participants yet? Thanks

Hi Team, We are following the documentation provided for the self-hosting egress service. I have a question about the websocket URL we are configuring for the livekit egress service. The WS URL we are passing in the client side(front-end) is the Livekit cloud URL and in our local self-hosting egress client, we are configuring the local WS URL. Is that the correct approach or does the WS URL need to be the same? We are trying to implement it with the LiveKit cloud but want to host the Egress Service locally.

slackbot spamming

Helllo All, i am building a RAG agent, i need to identify the user who asked the query, based on that I need to do the RAG, can anyone please help provide the code-snippet how to do this?

set the channel topic: This channel is being sunset, please ask any questions in the relevant channels rather than asking them here.

Hello #helpdesk I want to send some data to my livekit server from client side (next js) whenever new participant joins the room. data like key value pair ( chatbot id or etc ...) can anyone help me with this?

This message was deleted.

who do we need to speak to about issues with downloading invoices? I'm getting timeouts when trying to download PDF invoices