Open-source foundation voice models with 110ms latency and one-shot voice cloning for building expressive real-time voice agents.
Miso Labs builds highly emotive foundation models for voice with 110ms response latency—faster than human conversational latency—enabling fluid AI voice agents. The platform offers one-shot voice cloning from a 10-second audio clip and open-source models designed for local deployment, with on-premises hosting and enterprise support contracts available for data-sovereign deployments. Miso Labs is a product of Miso Labs.
For people
For agents
Nothing listed yet.