At the intersection of biopharma and artificial intelligence, a new star is rising at an astonishing pace. According to AIbase report, startup company Chai Discovery has recently announced a strategic partnership with pharmaceutical giant Eli Lilly (Eli Lilly). Under the agreement, Eli Lilly will adopt the company's AI software to accelerate drug development, focusing on using its algorithm called Chai-2 to develop antibody drugs.

This company, founded in 2024, has an impressive background. Its founder Josh Meier was once a core member of OpenAI, and Chai initially completed its startup incubation in OpenAI's office with support from CEO Sam Altman. In just one year, Chai not only secured a $130 million Series B funding round, but its valuation also soared to $1.3 billion, becoming one of the hottest names in the AI drug discovery field.

Chai Discovery aims to create a "computer-aided design (CAD) suite" for molecular design. Unlike traditional "shotgun" screening methods, the models developed by its laboratory focus on designing more efficient biological molecules from fundamental principles. This collaboration with Eli Lilly is seen as a watershed moment in the industry, with supporters believing that it signals that AI-discovered molecules will enter clinical trials on a large scale after 2026, thereby drastically reducing the time cost of drug development and overcoming previously unattainable medical challenges.

Key Points:

  • 🤝 Strong Partnership Between Giants: AI startup Chai Discovery has reached an agreement with pharmaceutical giant Eli Lilly to jointly design new antibody drugs using the Chai-2 algorithm.

  • 🦄 Backed by OpenAI: The founder worked at OpenAI, and the company was initially incubated in its office. It has already secured a $130 million Series B funding round with a valuation of $1.3 billion.

  • 🧪 Revolutionizing Traditional R&D: Through a customized model architecture (not open-source fine-tuning), it is dedicated to shifting drug development from random screening to precise "software-driven design."