Connecting Chai-3 to BMS's Antibody Discovery Engine

Bristol Myers Squibb (BMY) and private company Chai Discovery have entered a collaboration for therapeutic antibody discovery using generative artificial intelligence. This agreement is not a licensing deal for specific brands, generic names, or targets of clinical assets, but a discovery-phase program to design new antibodies for targets selected by BMS using the Chai-3 model. Therefore, the current value is first validated through early research metrics such as candidate generation speed, binding specificity, and development suitability, rather than clinical success rates. BMS can expand its experimental candidate pool using external models while reducing the time and cost of traditional screening.

The Key is Designing Novel Molecules, Not Existing Drugs

Chai-3 is a platform that computes protein structures and molecular interactions to design therapeutic antibodies and multispecific molecules, and the company provided performance data showing a doubling of antibody design success rates compared to the previous generation. Since the collaboration program is still before preclinical candidate selection, there are no direct clinical efficacy metrics to compare with Phase 1·2·3 or approved drugs. However, BMS has demonstrated capabilities in developing and commercializing antibody-based immuno-oncology therapies, such as Opdivo·nivolumab (PD-1), Yervoy·ipilimumab (CTLA-4), and Opdualag (PD-1·LAG-3). Opdivo and Yervoy received FDA approvals on December 22, 2014, and March 25, 2011, respectively, indicating a clear internal foundation to connect new platform outputs to clinical and regulatory stages.

Competition Occurs Simultaneously in AI Antibody Design and Internal Research at Large Pharma

Direct platform competitors include Generate Biomedicines, Isomorphic Labs, Absci (ABSI), AbCellera Biologics (ABCL), BigHat Biosciences, and Nabla Bio. In the commercial phase, Merck & Co. (MRK)'s Keytruda·pembrolizumab competes with Opdivo as a PD-1 standard therapy, with Keytruda first receiving FDA approval on September 4, 2014. The global monoclonal antibody therapeutics market is estimated at approximately USD 252.6 billion in 2024, meaning the success of a single candidate can create significant revenue opportunities, but the economic viability of platform agreements depends on clinical entry and asset ownership structures. Ultimately, the key metric for Chai to demonstrate is not the number of pharma partners, but the conversion rate of development candidates into clinical trial applications.

Financial Strength Extends Technology Validation Period

Chai Discovery raised USD 400 million in a Series C led by Index Ventures on July 14, 2026, with a post-money valuation set at USD 3.8 billion. With cumulative fundraising exceeding USD 600 million, the company now has greater capacity to invest in large-scale computing, wet-lab validation, and antibody engineering talent. The expansion of its customer base to include Eli Lilly (LLY), Pfizer (PFE), Novartis (NVS), and now BMS, supports the enterprise adoption of Chai-3. However, since this BMS collaboration does not involve a specific candidate, target, or clinical trial number, the next valuation reassessment factor will be the performance of preclinical candidates and clinical stage entries, rather than the number of research agreements.