Overview
Miso Labs builds advanced AI for voice applications. It develops foundation models that developers can use to create voice agents with warmth and personality, moving beyond purely functional chat. The core product is a set of voice-focused foundation models that deliver real-time responses with 110-millisecond latency, enabling natural, fluid conversations. It also offers one-shot voice cloning, letting a specific voice be replicated from a 10-second audio sample. For privacy-conscious enterprises, Miso Labs provides open-source models that can run on-premises, so voice data stays in-house under support contracts. Compared with others, its key differentiators are its emphasis on emotive, human-sounding voice, ultra-low latency, one-shot voice cloning capability, and on-prem/open-source options. The company’s goal is to help developers build voice agents that users perceive as more human, friendly, and engaging, with fast, private, and expressive AI interactions.
Work here?Claim your company
Simplify's Take
What believers are saying
- The June 3, 2026 MisoTTS launch gave Miso Labs immediate developer attention.
- YC lists Miso Labs active with a large seed round and hiring across research and GTM.
- The September 2026 careers page shows demand for inference, data, and partnerships talent.
What critics are saying
- Digiday reported creator backlash on June 12, 2026 after Tana Mongeau's clone video.
- Open-source voice cloning invites copyright, publicity-rights, and platform-policy fights immediately.
- Aden Teo said Miso can clone any voice; that invites an existential trust collapse.
What makes Miso Labs unique
- Miso Labs released Miso One and MisoTTS 8B on June 3, 2026.
- Their models claim 110-millisecond latency and ten-second voice cloning.
- YC Spring 2026 backing and open-source on-prem deployment separate them from hosted voice APIs.
Help us improve and share your feedback! Did you find this helpful?
Funding
Industry standards