Livekit

LiveKit provides open-source infrastructure and a cloud platform for developers building voice, video, and physical AI agents. LiveKit Agents handles speech processing, turn-taking, and model integration, while client SDKs and telephony services connect agents to web, mobile, and telephone users. Developers can build customer-support agents, multimodal assistants, and robotics applications using managed LiveKit Cloud or self-hosted infrastructure with the same APIs and core capabilities. Its broader platform includes inference routing and observability tools for examining conversations through session replays, traces, and transcripts.

Co-founders Russ d'Sa, CEO, and David Zhao, CTO, introduced LiveKit's original open-source audio and video infrastructure in 2021. Its WebRTC selective forwarding server coordinates media streams between users and agents. Beyond media transport, LiveKit develops turn detection that combines speech meaning with acoustic cues to identify when someone has finished speaking, without waiting for a transcript. Its open eot-bench suite and datasets evaluate the tradeoff between response latency and premature interruptions across 14 languages.

In 2026, the company reported adoption by over 300,000 developers and teams; in January, it reported more than one million LiveKit Agents downloads per month. That January, LiveKit announced a $100 million Series C led by Index Ventures at a $1 billion valuation.