LiveKit
A developer platform for building, testing, deploying, and scaling AI agents with real-time voice, video, and telephony integrations. Used by OpenAI, Nvidia, and Salesforce.
LiveKit is an open-source framework and developer platform designed for building, testing, deploying, scaling, and observing AI agents in production. It provides a complete stack for real-time communication and AI integration, enabling developers to create voice, video, and physical AI agents with minimal setup.
Key Features
- Open Source Framework: Build and test AI agents locally or in the cloud.
- Real-Time Communication: Supports voice, video, and telephony integrations.
- Inference Gateway: Access TTS, LLM, and STT models via a unified interface.
- Cloud Platform: Deploy and scale agents globally with LiveKit Cloud.
- Observability: Full-stack monitoring for every agent session.
- Multi-Model Support: Integrate with leading AI models like OpenAI, Google, and Cartesia.
- Telephony Integration: Enable agents to make or receive phone calls via SIP or phone numbers.
- Enterprise-Grade Infrastructure: GDPR, SOC 2 Type 2, and HIPAA compliant with 99.99% uptime.
Tech / Self-Hosting
LiveKit is open-source and can be self-hosted. The platform supports Kubernetes for scalable deployments and provides detailed documentation for local and cloud-based setups. For production use, LiveKit Cloud offers a managed solution with global infrastructure.