Latent Labs launches web-based AI model to democratize protein design

Latent Labs, about six months after emerging from stealth mode with initial funding, has launched LatentX, a web-based AI model designed to democratize protein design by enabling users to create novel proteins directly in their browser using natural language. According to CEO and founder Simon Kohl, who previously co-led DeepMind’s AlphaFold protein design team, LatentX has achieved state-of-the-art performance on various metrics, with a high percentage of its designed proteins proving viable in physical lab tests. Unlike AlphaFold, which predicts existing protein structures, LatentX can generate entirely new molecules such as nanobodies and antibodies with precise atomic structures, potentially accelerating therapeutic development. Latent Labs’ business model focuses on licensing LatentX to academic institutions, biotech startups, and pharmaceutical companies rather than developing proprietary medicines itself. While the foundational model is currently free to use, the company plans to introduce paid advanced features over time. This approach addresses the challenge that many organizations lack the resources to build their own AI models