Discussion about choosing the right LiveKit plan

Getting Started
1

Hi LiveKit team,

I’m a beginner building a production app called AmbientCare Maya. My app uses a LiveKit AI voice agent inside the application, and the agent is becoming a core part of the product. Users will talk to the agent frequently, and later we also want app-to-app voice/video calling, mobile notifications, and possibly telephony.

I’m considering the $50/month LiveKit plan. I want to understand whether this plan is enough for an early production launch.

My questions are:

  1. What exactly is included in the $50/month plan for AI voice agents?
  2. How many agent session minutes are included per month?
  3. How many WebRTC/audio/video participant minutes are included?
  4. What happens if I exceed the included usage — does the agent stop, throttle, or continue with overage billing?
  5. If many users are talking to the agent at the same time, what concurrency limits apply?
  6. Are AI model, STT, TTS, and inference costs included, or billed separately?
  7. Does the same plan cover my API usage if I’m using LiveKit through my app backend?
  8. For a beginner launching an app with many potential users, is the $50 plan enough, or should I consider a higher plan or self-hosting?
  9. How can I monitor usage and set alerts so my app does not unexpectedly stop or generate a very high bill?

My main concern is reliability: I do not want the agent to suddenly stop while users are using the app. Please explain the practical limits and recommended plan for an early production app.

2

Hi, I believe most of your specific questions should be answered by Pricing | LiveKit and Build voice, video, and physical AI | LiveKit but let me try and fill in the gaps.

Does the same plan cover my API usage if I’m using LiveKit through my app backend?

The APIs are included in the plan, with the exception of the Analytics API

For a beginner launching an app with many potential users, is the $50 plan enough, or should I consider a higher plan or self-hosting?

Yes, it is specifically designed for developers like yourself to launch their app into production. You would not need to upgrade to the Scale tier until the success and usage of your app warrants the additional quota included at that tier.

How can I monitor usage and set alerts so my app does not unexpectedly stop or generate a very high bill?

You can see your current usage at https://cloud.livekit.io/projects/p_/billing and you will receive email alerts

My main concern is reliability: I do not want the agent to suddenly stop while users are using the app.

The agent will not stop mid-conversation and you will not see the 429 quota exceeded messages etc, that you get on build tier when you exceed your quota.