Problem

Drug discovery is a slow, expensive, and failure-prone search through an enormous molecular design space. Bringing a new drug to market is commonly estimated to take roughly 10–15 years and cost more than $1–2 billion, while early-stage biologics research can require months of iterative laboratory work before researchers identify a viable candidate. Chai Discovery’s central use case is shortening that cycle, particularly for discovering drug-like antibodies against difficult therapeutic targets.

The clearest “killer” application is de novo antibody design: starting with a target or epitope, generating candidate molecules computationally, and testing them experimentally rather than relying on lengthy trial-and-error campaigns. Chai says its Chai-2 model achieves close to 20% antibody-design hit rates, with processes that previously required months potentially completed in weeks.

Product / Service

Chai builds a computer-aided design suite for molecules, using frontier AI models to predict and reprogram interactions between biochemical molecules. Its model family includes Chai-1, a multimodal foundation model for molecular-structure prediction, and Chai-2, a zero-shot antibody-design system. Researchers can specify the target, epitope, antibody format, and framework, then use the models to generate candidates with drug-like properties for experimental validation.

The delivery model is primarily enterprise deployment through pharmaceutical collaborations and licenses, rather than a conventional consumer software product. Chai provides access to its core models and platform, while also developing custom models trained on a partner’s proprietary data and tailored to that company’s workflows. The intended benefit is fewer wet-lab iterations, faster movement from target selection to candidate molecules, and the ability to pursue biological targets that conventional methods may not reach.

Market

Chai competes in generative-AI drug discovery, computational molecular design, and AI-enabled protein and antibody engineering. Its competitive set includes AI-native drug-discovery companies such as Isomorphic Labs, Recursion, EvolutionaryScale, and Xaira, as well as established computational-drug-design companies such as Schrödinger. Frontier-model companies including Anthropic and OpenAI are also cited as potential competitive pressure as general-purpose AI firms develop life-science offerings.

Chai is clearly beyond the pre-revenue experimentation stage in terms of commercial validation, although the available evidence does not disclose revenue or contract values. It has announced deployments or license and collaboration agreements with Eli Lilly, Pfizer, Novartis, and argenx; Lilly and Pfizer are using or planning to use Chai’s platform in real discovery workflows, including custom models. The company announced $400 million of new funding in July 2026, following earlier financing that had valued it at $1.3 billion, indicating substantial investor and pharmaceutical-industry traction.