💨 Abstract

Smallest.ai, a San Francisco-based AI company, launched Lightning, a fast real-time text-to-speech model that generates hyper-realistic audio in English and Hindi within 100ms. It reduces costs for voicebot builders, enabling population-scale use cases at sub 1 Rs/min. Lightning also plans to support other languages and can be used for creating audiobooks and voiceovers.

Courtesy: theprint.in