I'm happy to share that LiveKit can indeed host multiple AI agents with different numbers and scale them according to your business needs. The platform is designed to support a variety of AI agents, each with unique configurations and behaviors, and it offers built-in scalability through load balancing and elastic instance management to keep things running smoothly. When it comes to hosting multiple agents, you have the option of self-hosted deployments or using LiveKit Cloud. LiveKit Cloud is especially convenient because it provides automatic scaling and operational observability, which is fantastic for larger-scale deployments. The right plan for you will really depend on how many concurrent sessions you expect, with higher-tier plans supporting up to 50 concurrent sessions or even more. If you’d like, feel free to share more about your expected usage—like the number of agents and concurrent sessions—and I’d be more than happy to help recommend the best plan that fits your needs perfectly. Sources: LiveKit Pricing | Deployment and scaling | LiveKit Docs | Deployment and scaling | LiveKit Docs | README.md | examples/src/multi_agent.ts

I'm happy to help you get started! The Ship plan at $50 per month is a great choice if you're working with smaller-scale deployments, like hosting five clients. Here's what it offers: • Up to 20 concurrent agent sessions. • 5,000 agent session minutes per month (and if you need more, additional minutes are just $0.01 each). • 5,000 SIP participant minutes along with 150,000 WebRTC participant minutes included. • Supports up to 1,000 concurrent connections. For larger needs, the Scale plan at $500 per month is tailored to bigger applications, allowing up to 50 concurrent agent sessions with higher limits. Please feel free to reach out if you'd like more details or if I can assist you in getting started with the Ship plan for your initial testing. I'm here to help!