Chai Discovery

What Chai Discovery does

Chai Discovery is a private, venture-backed AI biotech founded in 2024 in San Francisco that builds multi-modal generative foundation models to predict and reprogram the interactions between biological molecules. Its core technology stack spans three generations of proprietary models — Chai-1 (multi-modal all-atom structure prediction across proteins, DNA, RNA, small molecules, and covalent modifications, released open-source under Apache 2.0), Chai-2 (zero-shot generative design of antibodies, nanobodies, and miniproteins achieving ~16% antibody binding and ~68% miniprotein wet-lab hit rates), and Chai-3 (next-generation model with doubled antibody-design success rates and multi-specific engineering, delivered via early-access channels). Wet-lab outcomes are independently validated, including Cryo-EM-confirmed sub-angstrom accuracy on full-length monoclonal antibodies designed against challenging GPCR and peptide-MHC targets.

The company monetizes primarily through multi-year strategic licensing agreements with large pharmaceutical companies, including a reported mid-eight-figure annual platform fee from Eli Lilly (signed Jan 2026) and a license with Pfizer (signed Jun 2026) that bundles early Chai-3 access with a custom model trained on the partner's proprietary data. A second tier of revenue comes from commercial subscriptions to Chai-2 via gated access, freemium academic access, and free-trial distribution of its miniprotein design suite to biotechs participating in Eli Lilly's TuneLab program. Go-to-motion is enterprise field sales into pharma R&D leadership; Chai also distributes through the NVIDIA BioNeMo ecosystem.

Name

Chai Discovery

Legal name

Chai Discovery

Website

https://chaidiscovery.com

Company type

Private

Founded year

2024

Operating status

Operating

Headcount range

11–50 employees

Short description

Chai Discovery builds proprietary AI foundation models (Chai-1, Chai-2, Chai-3) for biomolecular structure prediction and de novo antibody and miniprotein design, serving large pharmaceutical companies via multi-year licensing deals with Eli Lilly and Pfizer.

Ownership category

akta.pro rank

Product category

AI Drug Discovery

NAICS

Research and Development in Biotechnology (except Nanobiotechnology) (541714)

SIC

Services-Commercial Physical & Biological Research (8731)

akta.pro primary industry

AI/ML-Enabled Biologic Discovery Platforms (computational design, in silico screening) (HLAAAIAE)

akta.pro secondary industries

Drug Discovery (Target ID/Validation & Hit Discovery) (HLAIALAA), AI/ML Platforms for Drug Discovery & Experiment Optimization (HLAGAJAO)

Keywords

• AI drug discovery
• Protein design
• Antibody engineering
• Molecular structure prediction
• Biologics design

Headquarters

HQ city

San Francisco

HQ country

United States

HQ region

North America

Markets served

GTM type

B2B

Offering type

Software

Cost components

Personnel, Technology or R&D, Infrastructure, Marketing or Sales, Operations

Revenue model

1. Pharma licensing agreements: Multi-year strategic license agreements with pharmaceutical companies (Eli Lilly, Pfizer) granting access to foundation models and custom-trained models on the partner's proprietary data; upfront licensing value plus milestone economics.
2. Recurring annual platform fees: Multi-year, mid-eight-figure annual platform fees paid by large pharma for biologics design access, embedding the model into partner discovery workflows.
3. Commercial access subscriptions: Access to Chai-2 for commercial organizations via gated access request, with ongoing commercial relationships supporting research and product development.
4. Co-developed custom AI models: Revenue from developing purpose-built AI models trained on a partner's proprietary data, delivered as part of broader strategic collaboration packages (e.g., the bespoke model for Lilly, custom model for Pfizer).

Pricing tiers

Go-to-market motion1 record

Core offering

Chai Discovery develops multi-modal generative AI foundation models (Chai-1, Chai-2, Chai-3) that predict and reprogram biochemical molecule interactions. The company designs and engineers novel biologics—including monoclonal antibodies, nanobodies, scFvs, miniproteins, and other biomolecules—by predicting structures and generating de novo binders against specified targets. Its platforms serve pharmaceutical and biotechnology partners for drug discovery and biologics design.

Product overview

Chai Discovery offers a platform of proprietary AI foundation models for molecular design and drug discovery, comprising three generations of models: Chai-1 (multi-modal molecular structure prediction, publicly accessible), Chai-2 (zero-shot de novo antibody and miniprotein design, commercially licensed), and Chai-3 (next-generation model under early access). Chai-1 is distributed to the public via the free Chai Lab Web Interface (lab.chaidiscovery.com) and the open-source chai-lab software library on GitHub under Apache 2.0 license. Chai-2 and its Miniprotein Design Suite are accessed through commercial partnerships (e.g., Eli Lilly, Pfizer, Lilly TuneLab). Together the platform enables scientists to predict, design, and reprogram biomolecular interactions, compressing drug discovery timelines from months or years to weeks.

Differentiator

Problem solved

Functional benefit

Products and services

• Chai-1 Multi-modal foundation model for molecular structure prediction across proteins, small molecules, DNA, RNA, covalent modifications, and ligands. Achieves state-of-the-art accuracy on drug discovery benchmarks and runs single-sequence mode without MSAs while preserving most performance.
• Chai-2 Zero-shot generative platform for fully de novo antibody and miniprotein design, enabling epitope-specific binder generation without high-throughput screening. Achieves double-digit experimental success rates across scFvs, VHH/nanobodies, miniproteins, and full-length mAbs.
• Chai-2 mAb Extension of Chai-2 to full-length, drug-like monoclonal antibody design with developability characteristics on par with well-behaved therapeutic antibodies. Validated via Cryo-EM at sub-angstrom accuracy and demonstrated on challenging GPCR and peptide-MHC targets.
• Chai-3 Next-generation foundation model that doubles Chai-2's success rate in AI-driven antibody design, enables multi-specific molecule engineering, and reduces antibody design failure rates by half compared to Chai-2. Provided via early-access license to Pfizer.
• chai-lab (open-source library) Open-source software library providing Chai-1 model weights and inference code under Apache 2.0 license, enabling programmatic molecular structure prediction for free commercial and academic use.
• Chai Lab Web Interface Free web-based interface for running Chai-1 molecular structure prediction, available for commercial applications including drug discovery.
• Miniprotein Design Suite Chai's miniprotein design suite enabling in-silico molecular candidate generation without traditional library screening. Achieves extremely high success rates (>50%) for miniproteins.

Ideal customer profiles2 records

Technology focussed Yes

API detail

Has API

Yes

API docs

API detail

Core technology

AI maturity

App detail

Integration1 record

AI capability5 records

Feature5 records

Recent moves6 records

Expansion highlights3 records

Direct peers

• Isomorphic Labs: Alphabet-backed AI drug discovery company building foundation models for molecular structure prediction and drug design. Most direct competitor to Chai-1/2/3, targeting the same large-pharma licensing model.
• Insilico Medicine: AI-driven drug discovery company combining generative chemistry, target identification, and clinical development. Directly comparable AI-native biotech commercializing foundation models to pharma, with broader small-molecule focus but overlapping biologics ambitions.
• Generate Biomedicines: Flagship Pioneering–backed generative biology company building foundation models (Chroma) to program novel protein therapeutics. Closest direct competitor in generative biologics and antibody/miniprotein design.
• Absci: AI-driven generative biology company with an integrated wet lab for antibody design. Chai's co-founders include former Absci employees, and the company competes for the same pharma antibody-discovery budgets.
• Genesis Therapeutics: AI-native drug discovery company building foundation models (DynamicBPNet) for molecular simulation and protein-ligand interaction prediction. Targets the same pharma licensing and co-development model for AI-designed therapeutics.
• Iambic (formerly Entos): AI-driven drug discovery company building physics-based and ML foundation models for small-molecule and protein design. Comparable in stage, capital intensity, and pharma partnership strategy.

Broad incumbents

• Recursion Pharmaceuticals: Publicly listed AI-driven biotech combining high-throughput wet-lab experimentation with foundation models for target identification and small-molecule discovery. Operates at greater scale and broader scope than Chai, with established pharma partnerships.

Emerging players

• Cradle Bio: AI protein design startup focused on engineering novel protein therapeutics and enzymes. Comparable technology stack and target customer base (biologics-focused pharma), with somewhat narrower scope than Chai's full biomolecule platform.
• EvolutionaryScale: AI frontier-model company for biology, founded by former Meta FAIR researchers (overlap with Chai's founding team backgrounds). Competes in the foundation-model-for-biology layer that powers Chai-1/2/3.
• Aqemia: AI-driven drug discovery startup combining deep generative models with physics-inspired approaches for small-molecule discovery. Chai's co-founders include former Aqemia contributors; comparable in stage and AI-native biotech positioning.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat5 records

Key risks5 records

Key highlights6 records

Customer concentration

• GitHubgithub.com/chaidiscovery/chai-lab
• Xx.com/chaidiscovery
• LinkedInlinkedin.com/company/chaidiscovery

Revenue estimate

Valuation estimate

Number of profiles

Profiles7 records

Funding overview

Funding rounds6 records

Investors28 records

Funding detail is available on the Subscription and Enterprise plan.Contact sales →

M&A

Investments

M&A and investment is available on the Subscription and Enterprise plan.Contact sales →

Frequently asked questions about Chai Discovery