Smallest.ai

Smallest.ai develops a real-time voice AI platform built on small, efficient multi-modal models, offering speech-to-text, text-to-speech, speech-to-speech, and voice agent capabilities. Its product suite includes Lightning, a text-to-speech model with 100ms latency across 15+ languages; Pulse, a speech-to-text model supporting 38 languages with emotion and speaker detection; Electron, a sub-3B parameter language model that outperforms GPT-4.1; and Hydra, one of the first native speech-to-speech models built for production. The company philosophy centers on specialized small models that achieve capabilities once reserved for massive systems, offering faster reasoning, better efficiency, and deployment flexibility from edge devices to real-time applications. The agentic platform enables users to configure agents, select voices, set languages, and deploy from a single interface. Smallest.ai supports on-premises and private cloud deployments and holds ISO 27001, SOC 2 Type 2, GDPR, and HIPAA certifications. Founded in 2023 and headquartered in San Francisco, California, the company has raised $8.4 million in Seed funding from investors including Sierra Ventures, 3one4 Capital, Better Capital, LetsVenture, and the Development Bank of Wales.