Smallest.ai, a voice AI startup founded in late 2024, announced in July 2026 that it has raised $13 million in a Series A funding round, bringing its total funding to over $21 million. The round was led by Seligman Ventures, with participation from Sierra Ventures and 3one4 Capital.
The company is building a small, specialized voice model designed to make AI-powered phone calls indistinguishable from conversations with a human. Rather than relying on large language models (LLMs), Smallest.ai uses a compact model that processes audio in real time — listening, thinking, and responding simultaneously, much like a person would.
“While I’m speaking to you, you’re already thinking, and you might interrupt me if I talk for too long,” said Sudarshan Kamath, founder and CEO of Smallest.ai. “That’s exactly how our model is designed to work.”
The core technical challenge the startup addresses is latency. Standard LLMs require a full prompt before generating a response, creating pauses that feel unnatural in voice conversations. Smallest.ai’s model responds with virtually no lag for topics within its knowledge base. When a query falls outside that scope, it hands off to a larger foundational model — briefly placing the caller on hold, similar to how a human agent might pause to research an answer.
The model also handles diverse accents, supports dozens of languages, and is built to function in noisy environments. Current customers include RingCentral and Truecaller. Kamath identified customer support companies, including Sierra and Decagon, as potential future customers, arguing that building proprietary voice models would distract those firms from their core business.
Smallest.ai competes with ElevenLabs, Cartesia, and regional players such as Sarvam, but distinguishes itself by focusing exclusively on real-time conversational voice agents for enterprise use — not adjacent applications like audio dubbing or podcasting.
“We want our models to break the Turing test,” Kamath said. “You should speak to our model and not know it’s AI or human.”
Source: TechCrunch