What Genesis Molecular AI does

Genesis Molecular AI (formerly Genesis Therapeutics) is a private, venture-backed AI drug discovery company founded in 2019 by Evan Feinberg in South San Francisco. The company develops GEMS (Genesis Exploration of Molecular Space), an agentic AI operating system for small molecule drug discovery that integrates foundation models, autonomous agents, and exploratory tooling used by chemists across the full design-make-test cycle. GEMS is powered by Pearl, Genesis's flagship generative 3D foundation model for protein-ligand cofolding — the first model to clearly surpass AlphaFold 3 on public benchmarks (Runs N' Poses, PoseBusters, OpenBind) and on Genesis's internal crystal structures, achieving sub-angstrom accuracy via a SO(3)-equivariant diffusion architecture trained on physics-based synthetic data. The platform also includes DeCAF-Pearl, a flow map distilled variant delivering ~5x inference speedup, and predictive capabilities spanning potency, selectivity, and 30+ ADME properties.

The company operates a B2B enterprise partnership model with global pharmaceutical companies, structured as multi-target, multi-year collaboration agreements. Disclosed deal economics include a $35M upfront from Gilead (Sept 2024, three initial targets), $30M upfront from Incyte (Feb 2025, two oncology targets), and $120M upfront from Incyte (May 2026 expansion, including $40M equity), with up to $232M in milestones per program and over $1B in potential aggregate milestones across five initial targets. Genesis also advances a wholly-owned internal pipeline across oncology (pan-mutant PIK3CA inhibitors, programmed cell-death promoters) and immunology (validated and novel autoimmune targets), and operates three U.S. offices (San Mateo HQ + AI research, San Diego wet lab, New York AI research). The company has raised over $300M cumulatively across a $52M Series A (Rock Springs, Dec 2020) and a $200M oversubscribed Series B (a16z, Fidelity, BlackRock, NVIDIA NVentures, Aug 2023), and reports $16.1M in 2025 ARR per third-party data.

Name

Genesis Molecular AI

Legal name

Genesis Molecular AI, Inc.

Website

https://genesis.ml

Company type

Private

Founded year

2019

Operating status

Operating

Headcount range

101–250 employees

Short description

Genesis Molecular AI is a private AI drug discovery company, founded in 2019, that develops the GEMS platform and Pearl foundation model to accelerate small molecule discovery. It partners with major pharmaceutical companies (Incyte, Gilead) on multi-target collaborations and advances an internal pipeline in oncology and immunology.

Ownership category

akta.pro rank

Product category

AI Drug Discovery

NAICS

Research and Development in Biotechnology (except Nanobiotechnology) (541714), Software Publishers (5132)

SIC

Services-Prepackaged Software (7372), Services-Commercial Physical & Biological Research (8731)

akta.pro primary industry

AI/ML Platforms for Drug Discovery & Experiment Optimization (HLAGAJAO)

akta.pro secondary industry

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

Keywords

• AI drug discovery
• Foundation models
• Protein-ligand cofolding
• Small molecule design
• Computational chemistry

Headquarters

HQ city

South San Francisco

HQ country

United States

HQ region

North America

Offices3 records

Markets served

GTM type

B2B

Offering type

Software

Cost components

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

Revenue model

1. Partnership Upfront Cash Payments: Large upfront cash payments received as part of multi-year pharma collaboration deals. Incyte paid $80M upfront cash in the 2026 expanded deal; Gilead paid $35M upfront in 2024.
2. Equity Investments from Partners: Strategic partners purchase equity in Genesis as part of deal structures. Incyte made a $40M equity purchase in Genesis as part of the 2026 expanded agreement; NVIDIA also made an equity investment (November 2024).
3. Milestone Payments: Preclinical, clinical development, regulatory, and sales milestones up to $232M per program. Incyte deal could pay over $1B in milestones across the five initial targets, with several billion more possible with additional targets.
4. Royalties on Approved Products: Tiered royalties on sales of any approved collaboration products commercialized by partners.
5. Recurring Research Funding: Recurring research funding from Incyte to support AI model training and inference compute workloads as part of the expanded 2026 agreement.

Pricing tiers

Go-to-market motion2 records

Distribution channels2 records

Marketing channels8 records

Core offering

Genesis Molecular AI builds GEMS (Genesis Exploration of Molecular Space), an AI operating system for drug discovery that integrates proprietary foundation models, autonomous agents, and exploratory tooling used by chemists across the design-make-test cycle. The platform is powered by Pearl, a generative 3D foundation model for protein-ligand cofolding, and is deployed both into partners' R&D workflows via multi-target collaborations and into Genesis's own internal small molecule pipeline.

Product overview

Genesis Molecular AI offers a unified platform-plus-foundation-model product architecture. The flagship offering is GEMS (Genesis Exploration of Molecular Space), the AI operating system for drug discovery, which integrates foundation models, agents, and exploratory tooling for chemists. GEMS is powered by Pearl, Genesis' industry-leading foundation model for 3D biomolecular structure prediction that was the first model to clearly surpass AlphaFold 3 on protein-ligand cofolding. The architecture also includes DeCAF-Pearl (a distilled flow map variant achieving ~5x inference speedup) and the production Pearl System (which adds inference-time scaling and physics+AI-based pose ranking). GEMS powers both Genesis' internal wholly-owned pipeline (oncology programs including PIK3CA and programmed cell death, plus immunology programs) and major pharma partners' pipelines (Incyte multi-target collaboration in oncology, hematology, and inflammation; Gilead multi-target collaboration against hard-to-drug targets).

Differentiator

Problem solved

Functional benefit

Brands

• GEMS: Genesis Exploration of Molecular Space — the AI operating system/platform for drug discovery integrating foundation models, agents, and exploratory tooling for chemists.
• Pearl

Products and services

• GEMS (Genesis Exploration of Molecular Space) End-to-end agentic AI platform for small molecule drug discovery that integrates foundation models, autonomous agents, and exploratory chemistry tooling; used by pharma partners (Incyte, Gilead) and Genesis's internal discovery teams to generate, predict, and triage drug candidates.
• Pearl Foundation Model Generative 3D foundation model for protein-ligand cofolding and structure prediction built on a SO(3)-equivariant diffusion architecture trained on physics-based proprietary synthetic data; achieves sub-angstrom accuracy and powers GEMS for both internal pipeline and partner programs.
• DeCAF-Pearl (Denoiser Cofolding All-atom Flowmap model) First flow map model for all-atom cofolding, distilled from Pearl, providing ~5x faster inference at near-teacher accuracy to support high-throughput virtual screening and large-scale synthetic data generation for GEMS.
• Pearl System (Production Cofolding System) Production cofolding system combining the Pearl foundation model with inference-time scaling techniques and physics+AI-based pose ranking, deployed inside GEMS to support active drug discovery programs against hard-to-drug targets.
• Multi-Target Pharma Collaboration (Incyte) Bespoke multi-year, multi-target collaboration service where Genesis deploys the GEMS platform and forward-deployed drug hunters and engineers alongside partner R&D teams to co-develop differentiated small molecule drug candidates against partner-selected targets; includes exclusive development/commercialization rights, milestone payments, and royalties for the partner.
• Multi-Target Pharma Collaboration (Gilead) Bespoke multi-year, multi-target collaboration service where Genesis deploys the GEMS platform against hard-to-drug targets selected by Gilead; deal economics include upfront cash, milestone payments, and royalties on approved products.

Quantifiable outcome

• 15% relative improvement over AlphaFold 3 on Runs N' Poses protein-ligand cofolding benchmark
• +5 more outcomes

Named customers2 records

Segments3 records

Ideal customer profiles1 record

Technology focussed Yes

API detail

Has API

No

API docs

API detail

Core technology

AI maturity

App detail

Integration2 records

AI capability8 records

Feature12 records

Partnerships

Eight partnerships are on record, tiered flagship and minor.

• IncyteflagshipStrategic or Co-development Partner · 20 May 2026Significant expansion of the 2025 collaboration. Genesis receives $120M upfront ($80M cash + $40M equity) plus recurring research funding and is eligible for up to $232M in milestones per program, $1B+ in total milestones across five initial targets, and royalties. Incyte will provide proprietary experimental data to train GEMS — among the first major pharma-AI collaborations to power foundation model training with a partner's proprietary data. The companies will add at least five new collaboration targets (with the option to nominate more), and Incyte will have exclusive rights to develop and commercialize any resulting products.
• IncyteflagshipStrategic or Co-development Partner · 1 February 2025Initial 2025 strategic AI-focused research collaboration. Genesis received $30M upfront for two oncology targets using the GEMS AI platform. Genesis and Incyte are co-developing differentiated small molecule drug candidates against Incyte-selected targets, with Genesis providing AI-driven design and Incyte providing R&D and commercial capabilities.
• NVIDIAflagshipStrategic or Co-development Partner · 13 November 2024Equity investment and collaboration to optimize key computational methods for AI-powered drug discovery. Jointly announced Pearl at GTC 2025, with co-authorship of Pearl's technical report. NVIDIA's cuEquivariance kernels provide up to 15% training speedup and up to 80% inference speedup; collaboration accelerated Pearl inference 10–80% and unlocked 2.3–2.8x end-to-end speedup on B200 vs. baseline H100. NVIDIA's Triton and cuEquivariance were instrumental in accelerating Pearl inference.
• GileadflagshipStrategic or Co-development Partner · 10 September 2024Strategic collaboration to discover and develop novel therapies. Genesis received $35M upfront payment to collaborate on 3 initial hard-to-drug targets using the GEMS AI platform. The companies work closely together on discovery and development.
• OpenBind ConsortiumminorOthersOpenBind Consortium provided the first public structure-affinity benchmark for structure-based molecular AI; Genesis evaluated Pearl on OpenBind's EV-A71 2A 802-compound benchmark, achieving state-of-the-art zero-shot performance.
• Massachusetts Institute of Technology (MIT)minorStrategic or Co-development PartnerResearch collaboration with Ron Shprints, Peter Holderrieth, Juno Nam, Rafael Gómez-Bombarelli, and Tommi Jaakola on the DeCAF-Pearl work and related research.
• Carnegie Mellon UniversityminorStrategic or Co-development PartnerResearch collaboration with Nicholas Matthew Boffi on the DeCAF-Pearl work and related research.
• University College London / MilaminorStrategic or Co-development PartnerResearch collaboration with Joey Bose (UCL and Mila) on the DeCAF-Pearl work and related research.

Scale indicators13 records

Recent moves6 records

Expansion highlights6 records

Direct peers

• Recursion Pharmaceuticals: Public AI-driven drug discovery company combining high-throughput wet-lab experimentation with foundation models and partnering with major pharma (Roche/Genentech, Bayer, Sanofi). Comparable in scale, multi-target partnership model, and ambition to industrialize early-stage discovery.
• Isomorphic Labs: Alphabet/DeepMind spinout applying AlphaFold-derived foundation models to small molecule and biologics drug discovery, with multi-target pharma partnerships (e.g., Novartis, Eli Lilly). Directly comparable AI-first discovery platform with anchor pharma deals and proprietary structure-prediction models.
• Insitro: AI-driven drug discovery company combining ML, induced pluripotent stem cell biology, and high-throughput data generation, with multi-year partnerships with Bristol Myers Squibb, Gilead, and Eli Lilly. Closely comparable in business model, partnership economics, and biotech+ML integration thesis.
• Xaira Therapeutics: AI drug discovery company launched in 2024 with ~$1B from Sequoia and a re-purposed Illumina R&D engine, focused on foundation models for biology and small molecules. Comparable as a well-capitalized AI drug-discovery platform pursuing both partnerships and internal programs.
• Schrödinger, Inc. Public computational chemistry and AI drug discovery company offering physics-based simulation platforms and co-developing an internal and partnered pipeline. Comparable in combining physics-based and ML approaches and selling into pharma R&D, though more mature and platform-product-led.
• Generate Biomedicines: Flagship Pioneering-backed AI drug discovery company developing generative foundation models (e.g., Chroma) for proteins and increasingly small molecules, with major pharma partnerships (e.g., Novartis, Amgen). Closely comparable in generative AI foundation-model thesis and enterprise partnership model.
• Exscientia: Public AI-driven small molecule discovery company with end-to-end platform and pharma partnerships (e.g., Sanofi, Bristol Myers Squibb, Roche). Closely comparable in target class (small molecules), partnership economics, and platform-plus-internal-pipeline strategy.
• Atomwise: AI drug discovery company using structure-based and 3D molecular models (AtomNet) to discover small molecules, with a broad portfolio of pharma partnerships. Comparable in leveraging 3D structure prediction and partnership-led GTM, though earlier-stage and narrower on foundation-model capability.
• BenevolentAI: Public AI drug discovery company focused on target identification and small molecule design, with pharma collaborations (e.g., Merck KGaA, AstraZeneca). Comparable in AI-first small-molecule and target discovery positioning, though differentiated by an emphasis on target ID over cofolding.

Emerging players

• Iambic (formerly Entos): AI-driven small molecule discovery company combining physics-based simulation with generative AI, with NVIDIA collaboration and a partnership with AstraZeneca/Illumina. Closely comparable in NVIDIA partnership pattern and physics+ML foundation, though smaller in scale and fewer named anchor pharma partners.

Market position

Strengths5 records

Weaknesses5 records

Competitive moat6 records

Key risks7 records

Key highlights7 records

Customer concentration

• LinkedInlinkedin.com/company/genesistherapeutics
• Facebookfacebook.com/100067396426810

Revenue estimate

Valuation estimate

Number of profiles

Profiles13 records

Funding overview

Funding rounds6 records

Investors17 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 Genesis Molecular AI