Hi, We use the livekit react native SDK in our app. We've been experiencing issues with audio on android phones (specifically Samsung ones). These issues have been of two types: • The assistant interrupts itself, and its sound keeps getting fed back into the device. (prevalent on Galaxy S23 / S25 Ultra devices etc, and other Galaxy devices) • The voice input is not registered properly, and the one has to speak very loudly for the voice to get through. (we were able to recreate this on Galaxy S21 5G) We have Krisp disabled on the device, and use BVC. In our react native mobile app, our versions are:

"@livekit/react-native": "^2.9.1",
"@livekit/react-native-webrtc": "^125.0.12",

Our current audio options in Javascript are configured as:

*AudioSession*.*configureAudio*({ android: { audioTypeOptions: AndroidAudioTypePresets.media } });

We're trying other audio options now, if someone has recommended audio config specifically for android, let us know and we'll try that. If not, any other leads on this issue will be helpful!

Hi, I am trying to integrate a voice agent using livekit in my react-native mobile app. But the app is unable to establish connection with the agent. The agent is deployed in the backend and I am passing tokens in the frontend. The token generation is correct as I am able to connect with agent using the livekit starter examples. Can anyone help? Thanks!

publishing track {"enabled": true, "kind": "audio", "muted": false, "pID": undefined, "participant": undefined, "room": undefined, "roomID": undefined, "source": "microphone", "streamID": "ABCXXXX "streamTrackID": "ABCDYYYYY, "trackID": undefined}

SessionAPIUtilities.h:176 AudioSessionGetProperty (kMXSessionProperty_HasEchoCancelledInput) failed with error: -50 (sessionID: 0x0) I'm seeing this error while trying to setup livekit in my react native expo app (Ignite Boilerplate) Any suggestions ?

We have "react-native": "0.72.10",

Hi Team, There is an intermittent issue we are seeing. We are not disabling the camera or changing publishing permission, but the track is still being unpublished as soon as user published it. What could be the reason for this? Is there any issue going on with Livekit? Here are the logs for a session we captured. Please check. https://cloud.livekit.io/projects/p_2y5csewzsft/sessions/RM_4UxSJTQo92Uj/participants/PA_BFZaecDo69DB Thank You cc: @plain-wall-1315

Hey everyone 👋 We’re running into a critical issue in our Flutter app after updating

livekit_client

to 2.5.2 (which depends on

flutter_webrtc

1.2.0) — we had to update for Google Play 16KB page size compliance. Here’s what’s happening (both Android & iOS): • When the remote participant joins before the local one, both audio and video tracks of remote participant are

null

on the local side, even though the remote participant has already published them. • Audio can still be heard, but track objects remain

null

. • When the remote participant mutes/unmutes mic or toggles video, no track events are fired — it behaves like the tracks were never published. • When the local participant joins first, everything works fine. This didn’t occur with

livekit_client

2.4.8 and

flutter_webrtc

< 1.0. It’s a critical blocker for us — has anyone else run into this or found a workaround? 🙏 I see very similar issue on GitHub issues - but explanation isn't so good, so I needed to explain above, since GitHub issues are often ignored. https://github.com/livekit/client-sdk-flutter/issues/895

Hello everyone, I'm building a voice chat app using React Native (Expo) and LiveKit. My app was working perfectly a week ago, but now it's stuck on "connecting..." and never finishes. It's like it's frozen on the loading screen. Here's the confusing part: • My LiveKit server is working fine. I tested it with a different example app, and it connected without any problems. • This means the problem isn't with my server or my login keys. • I haven't changed any of my code since it was last working. What I see in my app: The logs show that it starts to connect, but it never succeeds. It doesn't even get to a point where it shows an "error" or "disconnected" message. It's just stuck.

import { LogLevel, setLogLevel } from 'livekit-client';

setLogLevel(LogLevel.debug);

Logs

room event connectionStateChanged {"args": ["disconnected"], "event": "connectionStateChanged", "pID": "", "participant": "", "room": "", "roomID": undefined}

I also tried listening manually using

room.on(RoomEvent,()=>{})

and getting only these logs

[LiveKit] Event: connectionStateChanged connecting

LOG  room event audioPlaybackChanged {"args": [], "event": "audioPlaybackChanged", "pID": "", "participant": "", "room": "", "roomID": undefined}

[LiveKit] Event: audioPlaybackChanged false

The

onError

in

<LiveKitRoom>

isn’t showing any errors. The main clue: I haven’t changed anything in the frontend or backend since it last worked. The example project connects to my server but my app doesn't. Has anyone faced a similar situation where your app gets stuck "connecting" even though your LiveKit server is proven to be working? Any ideas what could have changed or what I should check in my React Native project?

are we able to register a custom video track in React Native or are we locked in to only the camera or native screenshare?

Hello, everyone. I have been successfully using the Live Kit SDK for Unity WebGL for months now. A few days ago, however, I noticed than when I start a call like before (no code changes to the call mechanic for months) every user hears themselves with an echo. It doesn't matter if they are wearing headphones or just use the laptop mic. Each user hears themselves when speaking. Does anyone know anything about this or could suggest some ideas to try to fix it? Again, it used to work fine before and I have done no changes to the code. I am on the latest version of the SDK and always have been. There are no updates available. At least not in the docs. It is version 2.0.1

Windows has had some updates so it could be related to that

I need some help figuring out some stuff here. I just updated from version 3.2.3 to 3.5.1 for the flutter SDK. Everything was working fine with 3.2.3 but now when people join with iOS (iPhone 12/iOS 18.7.1) they sometimes doesn't see other participants because of room.remoteParticipants being empty (and in turn, our UI doesn't update properly) after i've connected. The funny part is, when i connect on iOS with the microphone muted everything works. Can someone help me figure out why users connecting with/without microphone on iOS effects the room.remoteParticipants list of values?? Edit: everyone connects and such where they can still hear each other. But since the UI is built based on whats inside the room.remoteParticipants they can't see each other. These are the values for room.remoteParticipants if i listen to all room changes. Again, why are these values null and why can other participants control them or am I doing something extremly weird in the way I handle my remote participants?

Connecting without microphone
flutter: videoTrackPublications.firstOrNull?.enabled: null
flutter: videoTrackPublications.firstOrNull?.muted: null
flutter: videoTrackPublications.firstOrNull?.track?.isActive: null

Connecting with microphone
flutter: videoTrackPublications.firstOrNull?.enabled: true
flutter: videoTrackPublications.firstOrNull?.muted: false
flutter: videoTrackPublications.firstOrNull?.track?.isActive: true

Edit: saw that this issue was already reported and working on

Hello there. I can see a lot of packages that expose Krisp functionality to Web, Swift, Android etc., but I was wondering if it's possible to enable it using the rust-sdks directly? Searching through the codebase didn't yield much in that regard, but maybe there are plans to add it there? Or, if the effort to add it is small, maybe a brief description on how, and then we could potentially add it ourselves

Hi Team Trying to build sample swift application from repo - https://github.com/livekit-examples/swift-example Getting these compilation errors. Any idea how to resolve these #C01LPRVTQMU #C07FVFGAUKX #C025KM0S1CK LiveKitExample-havcspdeirrqylgwxhmyyqjsuptx/SourcePackages/checkouts/components-swift/Sources/LiveKitComponents/Audio/AudioProcessor.swift519: Task-isolated value of type '() async -> ()' passed as a strongly transferred parameter; later accesses could race

with livekit react native Screenshare example: https://github.com/livekit/client-sdk-react-native?tab=readme-ov-file#ios-1 https://jitsi.github.io/handbook/docs/dev-guide/dev-guide-ios-sdk/#screen-sharing-integration We got the system more or less working but are running into an issue when the user opens the broadcast picker but then closes it without selecting to start the broadcast. When we lanch the broadcast picker, and then dont select a broadcast, we are unable to kill the screen_share track generated from setting screenShareEnabled(true). When we select the broadcast and run it we have no issue killing the screenshare track. Are there any docs showing how to handle this? It doesnt look like there is a callback or anything that lets us know the picker was closed without an action.

https://github.com/livekit/components-swift/issues/30 Can someone help with this?

🆘 Hello LiveKit team! I’ve opened a PR for our Spitch plugin [tts/stt]. Could you review it? can someone help review it: https://github.com/livekit/agents/pull/3748

🔴 Hi LiveKit Team i am trying to import LiveKit in UiKit xcode26 i am having an error saying

Missing required module 'LKObjCHelpers' error when importing LiveKit

can anyone please help me fix this. i am using spm.

Can anyone confirm a working install for LiveKit with Unity? The default installation instructions are not working for me. I posted an issue on that too: https://github.com/livekit/client-sdk-unity/issues/166

To be clear, I'm also trying version 1.3 currently. The last version that worked for me was 1.2.5 so I'm now trying to troubleshoot across versions and device types. Right now, I can't get 1.3 to work with a Unity build for Android on Mac. Previously on 1.2.5, I was able to do that on Windows. So I'll be investigating a few variants.

hello, I'm developing an AI voice calling app using LiveKit in React Native (Expo). The app works perfectly when switching to the home screen or other apps like Notes, and even during lock screen — the call continues without any issues. However, on iOS, the call gets disconnected when I open apps such as Snapchat or WhatsApp (which use the microphone). This issue is already fixed on Android, but I’m looking for a solution or guidance on how to keep the call active on iOS even when switching to these apps. Can anyone help with this?

Hey all, I had a question about implementing a client SDK in Java for server-server communication. I have successfully implemented the

Join

protocol. I can see a correct

JoinResponse

followed by a room

Update

and then an

Offer

with various

Trickle

candidates coming in. If I then use the server SDK to query the room participants, I see my connected client no problems. I then start sending pings every 3 seconds. About 20 seconds into the entire process, I then receive a websocket disconnection (code 1000, reason = null) and my connected client is removed from the livekit room. I have been staring at other sdks for over 5 hours now and am about to crash out, does anyone have any idea why livekit is disconnecting me remotely? Edit: We are following everything documented here (https://docs.livekit.io/reference/internals/client-protocol/#joining-a-room).

Hi everyone, I have a question. I managed to set up a call via LiveKit, but now I need to make it work with other audio sources, like in-game sound or video playback. LiveKit has an AudioManager that handles the audio, and I don't want to disable it—I'd like to extend its functionality. Are there any best practices for this kind of case?

We’re currently facing a critical issue with our iOS app — the outgoing audio call disconnects or gets suspended when the app goes into the background. Our goal is to perform outgoing audio calls (no video) using LiveKit, where we create a room using a server URL and token. The call works perfectly while the app is active, but once it’s backgrounded, it stops. We’ve already: • Enabled Background Modes (

audio

,

voip

) in both Xcode and app.json / Info.plist • Called

setupCallService()

before

registerGlobals

in

App.tsx

• Started the native audio session before

startCallService()

• Verified our

AVAudioSession

configuration Despite all these configurations, iOS still suspends the app when it’s backgrounded. We couldn’t find any proper or complete guide to make LiveKit outgoing audio calls persist in the background on iOS. We urgently need your help and detailed guidance or configuration steps to fix this issue. we need help, the docs not helpful

Hello! Android compose library user here. I am using the latest compose livekit library

1.4.0

from july. The core library seems to be getting updates quite frequently. The

1.4.0

Compose lib is still using 2.18.3 of the core lib. Atleast that's the version I get as transitive dependency. Could you update the compose to use the latest core one to get all fixes available to compose users 🙂

Hello, we are trying to use the noise reduction and echo cancellation features of LiveKit (not the AI ones) in our Flutter application. The target platform is Linux. Our configuration looks like this:

RoomOptions(

defaultAudioOutputOptions: AudioOutputOptions(speakerOn: true),

defaultAudioCaptureOptions: AudioCaptureOptions(

echoCancellation: true,

noiseSuppression: true,

),

),

It seems both echo cancellation and noise suppression do not seem to have any effect during calls. Could you please confirm if these features are currently supported on Linux Flutter, and if any additional setup or dependencies are required to make them work? Thank you 🙂

hey we're running into a strange issue with livekit's ios sdk and an audio track from a separate source. repro steps: 1. separate source audio coming though speaker 2. start livekit conversation (coming though speaker) 3. go back to separate audio source (this is now coming out of the in ear speaker) any ideas on why this would be the case?

Hello Community, We’re integrating outgoing-only audio calls using the following SDKs in our React Native Expo app:

@livekit/react-native-expo-plugin

,

@config-plugins/react-native-webrtc

,

@livekit/react-native

,

@supersami/rn-foreground-service

, and

react-native-callkeep

. Everything works fine in the foreground, but when the app goes into the background, the audio call disconnects after about 1 minute. We have already: • Enabled

UIBackgroundModes

→

audio

and

voip

in Info.plist and

app.json

• Set up

registerGlobals()

and

setupService()

in App.tsx • Given microphone and background permissions • Not implemented PushKit (since we only do outgoing calls) Our main question: Do we need to add the VoIP entitlement certificate from Apple Developer Portal even though the app only handles outgoing calls? Or is there any additional iOS configuration, permission, or setup (in Xcode,

Info.plist

,

app.json

, or entitlement files) required to keep an outgoing call active in background mode? If anyone can share references, configuration examples, or a working setup for this scenario, it would be really helpful.

Hello, I couldn’t find this in the docs, but here’s my question: In the native SDK (iOS specifically) I see that lib directly taps into the microphone. Is there any way to pass to the SDK a stream of audio buffers that will be in turn passed to WebRTC? And if not, is there any way the SDK can feed us the buffers that it records so we can use them for things like VAD, volume, sound detection etc.? Thanks 🙂

Hello, I am trying to implement Picture in Picture, using Livekit Flutter Sdk. Android works fine, but when I minimize the app, camera stops working and I am seeing black screen on PiP. I didn’t find solution, what can be a reason, or does it work in PiP iOS at all?