Chai Discovery

Chai Discovery is an American artificial intelligence company headquartered in San Francisco, focused on building generative biology models for molecular design and protein-structure prediction. The company was founded in 2024 by Joshua Meier, Jack Dent, Matthew McPartlon, and Jacques Boitreaud, and develops Chai-1 (the open-source structure-prediction model released September 2024) and Chai-2 (the zero-shot antibody-design platform released in 2025). As of May 2026, Chai Discovery is one of the principal commercial AI-for-biology startups, with a $130 million Series B in December 2025 at a $1.3 billion valuation co-led by Oak HC/FT and General Catalyst, and an announced collaboration with Eli Lilly on biologics discovery.

At a glance

• Founded: March 2024 in San Francisco, California, by Joshua Meier, Jack Dent, Matthew McPartlon, and Jacques Boitreaud.
• Status: Private. Series B in December 2025 at a $1.3 billion valuation.
• Funding: More than $225 million cumulative across three disclosed rounds. $30 million seed (September 2024) led by Thrive Capital, OpenAI, and Dimension. $70 million Series A (August 2025). $130 million Series B (December 2025) co-led by Oak HC/FT and General Catalyst at a $1.3 billion post-money valuation.
• CEO: Joshua Meier, Co-Founder and Chief Executive Officer. Former OpenAI research and engineering staff.
• Other notable leadership: Jack Dent, Co-Founder. Former Stripe engineer and Harvard classmate of Meier. Matthew McPartlon, Co-Founder. Jacques Boitreaud, Co-Founder.
• Open weights: Partial. Chai-1 was released open-source for non-commercial research use; Chai-2 is closed and accessible through commercial partnerships.
• Flagship products: Chai-1 (multi-modal structure-prediction model, September 2024); Chai-2 (zero-shot antibody-design platform, 2025); custom Eli Lilly model under the January 2026 collaboration agreement.

Origins

Chai Discovery was founded in March 2024 by Joshua Meier and Jack Dent, both Harvard computer-science classmates who later took separate paths through the technology industry. Meier joined OpenAI in 2018 as a research and engineering team member during the lab's pre-GPT-3 period, then worked at biotech firm Absci on protein-design applications. Dent went to Stripe as a software engineer. The pair were joined by Matthew McPartlon and Jacques Boitreaud, both AI researchers with prior backgrounds in structural biology and machine learning.

The founding thesis traced back to a conversation with Sam Altman approximately six years before the company's launch, when Altman proposed to Meier and Dent the idea of applying generative AI to proteomics. The thesis became more tractable as protein-language-model and structure-prediction approaches advanced through 2022 and 2023, with AlphaFold 2 and the ESM protein-language-model line establishing that AI systems could match laboratory-style accuracy on protein-structure problems. Chai Discovery was incorporated as a venture-backed company in March 2024 with initial office space provided by OpenAI in San Francisco's Mission District.

The September 2024 release of Chai-1 was the company's first major public-facing milestone. Chai-1 was a multi-modal foundation model for molecular structure prediction, released open-source under a non-commercial research license. It positioned the company alongside Isomorphic Labs and EvolutionaryScale as one of the principal AI-for-biology research organizations producing flagship structure-prediction systems.

The 2025 release of Chai-2 marked the company's transition from open-source structure prediction toward closed-source generative-design products. Chai-2 was characterized in the company's announcements and in industry coverage as the first zero-shot generative platform to achieve double-digit experimental hit rates in de novo antibody design, with reported success rates of approximately 20 percent across 52 designed targets. The company's framing of this milestone was that an antibody-design cycle that traditionally requires 12 to 24 months could be compressed to four to eight weeks.

Mission and strategy

Chai Discovery's stated mission is to transform biology from science into engineering. The strategic premise is that frontier-scale AI applied to biological molecules can compress drug-discovery timelines from years to weeks, and that the resulting capability advantage justifies investment in the laboratory and modeling infrastructure needed to support it.

The strategy combines three threads. First, foundational research on generative biology models, with the Chai-1 and Chai-2 lines providing both the technical base and a research-credibility flywheel. Second, custom partnerships with pharmaceutical companies under which Chai trains purpose-built variants of its platform on partner-specific data and biological targets. Third, internal molecule-design programs that the company can advance through the discovery pipeline either alone or with partners.

The competitive premise reflects Chai Discovery's positioning at the intersection of generative-AI research talent (the OpenAI and Stripe origin), biology-specific application focus (antibody design as the lead application), and structural-prediction research history (the Chai-1 release as the public-credibility anchor). Industry coverage has consistently characterized the team as one of the more technically credible commercial AI-for-biology efforts, distinguished from peer biology-AI startups by the open-source release of Chai-1 and the experimental-hit-rate benchmarks reported for Chai-2.

Distribution channels include direct pharmaceutical-company partnerships (Eli Lilly), cloud and API delivery for selected biotech customers, and continued open-source research releases for non-commercial use.

Pipeline and programs

• Chai-1. Multi-modal foundation model for molecular-structure prediction, released September 2024. Open-source under non-commercial research license. Positioned as one of the principal AlphaFold-class structure-prediction systems globally.
• Chai-2. Zero-shot antibody-design platform released in 2025. Closed-source. Reported approximately 20 percent experimental hit rates across 52 de novo designed targets, characterized in coverage as a 200x improvement over prior computational methods.
• Eli Lilly custom model. Announced January 2026. Chai will develop a purpose-built AI model exclusively for Lilly, trained on Lilly's proprietary data and tailored to Lilly's biologics-discovery workflows. The collaboration covers multiple targets across Lilly's biologics portfolio.

Distribution channels include direct pharmaceutical partnerships, the open-source Chai-1 release, and continued model-iteration cycles through 2026.

Benchmarks and standing

Chai Discovery's evaluation framework centers on protein-structure-prediction accuracy (CASP-style assessments), antibody-design experimental-hit rates, and published research output. Chai-1 has been characterized in AI-for-biology coverage as competitive with leading structure-prediction systems including AlphaFold 3 from Isomorphic Labs and the ESM3 model from EvolutionaryScale.

Chai-2's reported approximately 20 percent experimental-hit rate on 52 zero-shot de novo antibody-design tasks is the most commonly cited capability benchmark associated with the company. Industry coverage has characterized the result as one of the strongest published zero-shot antibody-design hit rates in the commercial AI-for-biology category, though direct head-to-head comparisons with peer commercial platforms remain limited because most competitors do not publish equivalent benchmarks.

The company has not entered standard horizontal-LLM benchmarks (Artificial Analysis Intelligence Index, LMArena, GPQA Diamond) because biology-specific structure-prediction and antibody-design evaluations diverge from the LLM-leaderboard frame.

Leadership

As of May 2026, Chai Discovery's senior leadership includes:

• Joshua Meier, Co-Founder and Chief Executive Officer. Harvard computer-science alumnus, former OpenAI research and engineering staff (2018), and former machine-learning lead at biotech firm Absci.
• Jack Dent, Co-Founder. Harvard classmate of Meier, former Stripe software engineer.
• Matthew McPartlon, Co-Founder. AI researcher with prior background in structural-biology machine learning.
• Jacques Boitreaud, Co-Founder. AI researcher with prior background in protein modeling.
• Senior research and engineering leadership across the Chai-1, Chai-2, and partner-program lines.

The team's hybrid composition (frontier-AI research and engineering credentials combined with structural-biology domain experience) has been characterized in industry coverage as one of the company's principal recruiting and credibility advantages.

Funding and backers

• Seed (September 2024): $30 million led by Thrive Capital with OpenAI and Dimension participating.
• Series A (August 2025): $70 million.
• Series B (December 2025): $130 million co-led by Oak HC/FT and General Catalyst, at a $1.3 billion post-money valuation. Other participants include Menlo Ventures, OpenAI, Dimension, Thrive Capital, Neo, Yosemite venture fund, Lachy Groom, SV Angel, and new investors Glade Brook and Emerson Collective.

Cumulative capital exceeds $225 million across the three rounds. The continued OpenAI participation across rounds is a structural feature: the lab provided the company's first office space in 2024 and has remained an investor through Series B.

Industry position

Chai Discovery occupies a distinctive position among 2024-vintage AI-for-biology insurgent labs. The combination of OpenAI-affiliated founder credentials, the open-source Chai-1 release that established research credibility, the Chai-2 zero-shot antibody-design results, and the Eli Lilly partnership signal a company with both technical depth and a clear commercial channel into pharmaceutical R&D budgets.

Strategic risks include the open question of whether reported zero-shot hit rates on initial antibody targets generalize to broader therapeutic-target classes, the competitive pressure from Isomorphic Labs (which has the AlphaFold lineage and Alphabet-scale resources), and the dependence on a small number of large pharmaceutical partnerships for commercial revenue at this stage.

Strategic strengths include the technical credibility of the founding team, the open-source Chai-1 release that anchored research positioning, the Chai-2 experimental benchmarks that distinguish the company from peers without published hit rates, and the Eli Lilly collaboration that establishes a major-pharma reference customer.

Competitive landscape

Chai Discovery competes with several AI-for-biology peers across both research and commercial dimensions:

• Isomorphic Labs. The Alphabet-owned AI-drug-discovery company built on the AlphaFold lineage. Direct technical peer in structure prediction and generative biology.
• EvolutionaryScale. Protein-language-model lab from former Meta AI / FAIR ESM team members. Overlapping research direction in protein structure and design.
• Inceptive. RNA-foundation-model startup. Adjacent generative-biology positioning with different molecular focus.
• Lila Sciences. Autonomous-laboratory AI-for-science peer with broader research scope.
• FutureHouse. Nonprofit AI-for-science peer; agent-based research focus rather than direct molecular-design.
• Recursion Pharmaceuticals, Insitro, Generate Biomedicines. Public and late-stage private AI-drug-discovery peers with portfolio approaches.
• AbCellera, BigHat Biosciences, Profluent. Commercial peers in antibody-discovery and protein-design tooling.

Outlook

• The cadence of new model releases beyond Chai-2 across the 2026 calendar.
• The progression of the Eli Lilly collaboration through experimental-validation milestones and into clinical-stage molecule pipelines.
• The expansion of pharmaceutical-partnership reference customers beyond Lilly through 2026 and 2027.
• The continued tension between Chai Discovery's open-source research positioning (Chai-1) and its closed commercial product strategy (Chai-2).
• The trajectory of valuation across any 2026 to 2027 follow-on financing rounds.
• The competitive dynamics with Isomorphic Labs and other large-incumbent AI-drug-discovery efforts.

Sources

• Chai Discovery official site. Company reference.
• Chai Discovery Series B announcement. December 2025 funding announcement.
• Eli Lilly collaboration announcement. January 2026 partnership announcement.
• TechCrunch: From OpenAI's offices to a deal with Eli Lilly. Founder backstory and partnership context.
• TechCrunch: OpenAI-backed biotech firm Chai Discovery raises $130M Series B. Series B funding context.
• Bloomberg: Chai Discovery Gets OpenAI, Thrive Capital Backing. September 2024 seed-round context.
• Contrary Research: Chai Discovery Business Breakdown. Company analysis.