I have one more question regarding our platform architecture and pricing. We are planning to use one LiveKit account/project for our entire platform. Under this single account/project, we may have multiple hospitals, and each hospital would have its own

Getting Started
1

Hi LiveKit Team,

I have one more question regarding our platform architecture and pricing.

We are planning to use one LiveKit account/project for our entire platform. Under this single account/project, we may have multiple hospitals, and each hospital would have its own AI voice agent.

For example:

One LiveKit Account / Project
│
├── Hospital A
│   └── Agent A
│       ├── Phone Number 1
│       ├── Phone Number 2
│       └── Phone Number 3
│
├── Hospital B
│   └── Agent B
│       ├── Phone Number 4
│       └── Phone Number 5
│
└── Hospital C
    └── Agent C
        ├── Phone Number 6

So, each agent will have at least one phone number, but some agents/hospitals may need multiple phone numbers.

We would like to understand the pricing and limits for this setup.

Specifically:

  1. Can we have multiple AI agents under one LiveKit account/project?

  2. Can each agent have one or multiple phone numbers?

  3. Is there any additional LiveKit cost for creating/deploying multiple agents under the same account/project?

  4. Is there any additional cost per phone number/DID, and how is telephony pricing calculated?

  5. If we have, for example, 10 agents and 20–30 phone numbers under one LiveKit project, would there be any additional platform, agent, dispatch-rule, trunk, or telephony charges apart from the normal usage-based costs?

  6. Are there any limits on the number of agents, phone numbers, trunks, or dispatch rules that can be created under a single project?

Our goal is to build a multi-tenant healthcare platform where one LiveKit account/project manages multiple hospital-specific agents and their phone numbers, rather than creating a separate LiveKit account/project for every hospital.

Could you please clarify the pricing model and recommended setup for this architecture?

Best regards,
Lily
Quto AI

2

I think this answers all those questions

Let me know if something is not clear there.

3

@Lily_Singhal, One project with many agents is the intended model, and billing is usage-based, so the multi-tenant structure itself carries no extra fee.

  1. Multiple agents per project: yes. You are metered on concurrent agent sessions across the project, not the number of agents, at $0.0100/min (pricing).
  2. Numbers per agent: one or many, mapped through each agent’s trunk and dispatch rule, as in your other thread.
  3. Deploying more agents: no per-agent or platform fee. You pay session minutes and concurrency only.
  4. Per number: LiveKit Phone Numbers carry a monthly rental (Build includes 1 US local number and 50 inbound minutes) plus $0.0100/min telephony (quotas & limits). A third-party trunk (Twilio, Telnyx) bills the DID on their side, and LiveKit bills only SIP minutes at $0.02/min. For 20-30 numbers, BYO trunk usually keeps DID rental with your carrier.
  5. Dispatch rules and trunks are configuration objects, not separately billed.
  6. Limits: the metered quotas are concurrency, number rentals, egress time, and observability, not a cap on how many agents, trunks, or dispatch rules you create. Increases are requestable on that same quotas page.

Only the Ship, Scale, and Custom allotments, and a limit raise for a 20-30-number healthcare rollout, need LiveKit sales.