Smallest.ai raises $13M to make voice AI pass the Turing test
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Smallest.ai raised $13 million in a Series A led by Seligman Ventures, with Sierra Ventures and 3one4 Capital participating, bringing total funding to more than $21 million
- Sudarshan Kamath, Smallest.ai's founder and CEO, said the company's voice model is designed to listen, think, and speak simultaneously, mimicking how humans handle conversation flow and interruptions
- Smallest.ai's model operates as a real-time intelligence layer with near-zero response lag; when queries fall outside its scope, it hands them off to a large foundational model that briefly 'researches' the answer
- RingCentral and Truecaller are named as existing customers, while AI support startups Sierra and Decagon are cited as potential customers
- Kamath argued that well-funded customer support companies won't build their own voice models, calling voice excellence 'a distraction from their core business'
- Smallest.ai competes with ElevenLabs, Cartesia, and regional players like Sarvam, and unlike some rivals focuses strictly on real-time conversational agents rather than audio dubbing or podcasting
- The company's stated mission is to 'break the Turing test' — making voice AI indistinguishable from a human speaker
Why it matters: If Smallest.ai's bet on small, specialized voice models proves out, it could become a default infrastructure layer that every AI customer-support platform plugs into, siphoning a high-leverage slice of the enterprise voice stack from competitors like ElevenLabs while making it harder for support startups to justify building voice in-house.
