Valence Discovery
Valence Labs is an AI-first drug discovery subsidiary of Recursion, founded 2018 in Montréal, that builds foundation models for biology and chemistry — including TxPert and the Virtual Cells framework — using Recursion's BioHive supercomputer and >60 petabyte proprietary dataset to target pharmaceutical and biotechnology applications.
- Company typePrivate
- Founded2018
- HeadquartersMontréal, Canada
- Headcount11–50
- GTM typeB2B
- OfferingSoftware
What Valence Discovery does
Valence Discovery, operating publicly as Valence Labs, is an AI-first drug discovery company founded in 2018 in Montréal, Canada. It designs novel chemistry against previously intractable biological targets by combining machine learning with chemistry and biology, building foundation models that operate on Recursion's proprietary dataset of more than 60 petabytes of experimental biological data and on the BioHive supercomputer.
The company's core technical outputs are three: (1) the Virtual Cells framework, a long-term research program aimed at computational whole-cell modeling; (2) TxPert, a foundation model launched in May 2025 that predicts cellular responses to perturbations across modalities; and (3) OpenQDC, an open-source benchmark and leaderboard for molecular machine learning models released in November 2024. The team is small (11–50 employees) and led by co-founders Daniel Cohen (CEO), Therence Bois (COO), and Prudencio Tossou.
Valence operates as a wholly-owned subsidiary of Recursion and functions as Recursion's dedicated AI research engine. No independent external revenue, pricing model, or named external customer is disclosed in available material; the stated target market is pharmaceutical and biotechnology companies. Cumulative independent funding prior to and around the Recursion acquisition totaled approximately $8.5M, with the largest round being a $7M financing in December 2021 led by Air Street Capital. The economics of the business are effectively bound to Recursion's internal R&D priorities and platform.
Valence Discovery firmographics
Firmographics- Name
- Valence Discovery
- Legal name
- Valence Labs
- Website
- https://valencediscovery.com
- Company type
- Private
- Founded year
- 2018
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- Valence Labs is an AI-first drug discovery subsidiary of Recursion, founded 2018 in Montréal, that builds foundation models for biology and chemistry — including TxPert and the Virtual Cells framework — using Recursion's BioHive supercomputer and >60 petabyte proprietary dataset to target pharmaceutical and biotechnology applications.
- Ownership category
- akta.pro rank
Valence Discovery industry classification
Industry- Product category
- AI Drug Discovery Platform
- NAICS
- Research and Development in the Physical, Engineering, and Life Sciences (54171), Research and Development in Biotechnology (except Nanobiotechnology) (541714)
- SIC
- Services-Commercial Physical & Biological Research (8731), Services-Prepackaged Software (7372)
- akta.pro primary industry
- AI/ML Platforms for Drug Discovery & Experiment Optimization (HLAGAJAO)
- akta.pro secondary industry
- Cheminformatics & Molecular Modeling / CADD (HLAGAJAE)
Keywords
Where Valence Discovery is headquartered
LocationHeadquarters
- HQ city
- Montréal
- HQ country
- Canada
- HQ region
- North America
Markets served
Valence Discovery business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Personnel, Technology or R&D, Infrastructure, Operations
Marketing channels3 records
Valence Discovery product offering
Product offeringCore offering
Valence Labs develops AI-first computational models—called virtual cells—that predict the functional response of human cells to genetic and chemical perturbations, enabling pharmaceutical and biotech companies to generate, test, and refine therapeutic hypotheses before clinical trials. The platform is built on Recursion's >60 petabytes of interventional biological data and BioHive supercomputer, and includes flagship products such as the TxPert predictive model and the OpenQDC open-source quantum mechanics dataset hub.
Product overview
Valence Labs is Recursion's AI research engine that develops computational approaches to predict, explain, and decode biology for drug discovery. The core offering is Virtual Cells—mechanistic models of cellular function that predict cellular responses to interventions. Key products include TxPert, a state-of-the-art model for predicting transcriptional responses to genetic perturbations, and OpenQDC, an open-source hub of ML-ready quantum mechanics datasets. These products are powered by Recursion's data generation infrastructure (>60 petabytes) and BioHive supercomputing resources.
Differentiator
Problem solved
Functional benefit
Products and services
- Virtual Cells Mechanistic computational models of cellular function that accurately predict the functional response of cells to genetic and chemical interventions, enabling pharmaceutical and biotech researchers to generate, test, and optimize therapeutic hypotheses before human clinical trials.
- TxPert A state-of-the-art predictive model leveraging multiple biological knowledge networks to accurately predict transcriptional responses under out-of-distribution scenarios for unseen genetic perturbations, designed for AI-driven drug discovery researchers.
- OpenQDC Open-source hub of ML-ready quantum mechanics datasets consolidating 40+ QM datasets covering 1.5 billion geometries across 70 atom species and 250+ QM datasets, intended for ML researchers and practitioners in computational chemistry and drug discovery.
Companies that use Valence Discovery
Customer profileSegments1 record
Ideal customer profiles1 record
Valence Discovery technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability7 records
Feature5 records
Valence Discovery partnerships and signals
Strategic signalScale indicators1 record
Recent moves6 records
Expansion highlights4 records
Valence Discovery competitors and assessment
Company assessmentDirect peers
- Recursion Pharmaceuticals: Valence Labs' parent company and direct peer in AI-driven drug discovery. Recursion operates the BioHive supercomputer, automated wet-lab infrastructure, and >60 PB of interventional biology data that power Valence's Virtual Cells platform. After merging with Exscientia, Recursion is the most directly comparable full-stack AI drug discovery company.
- Insitro: Insitro is a direct competitor in ML-driven drug discovery, combining induced pluripotent stem cell-derived data with machine learning models to predict therapeutic outcomes. Like Valence, it pursues predictive biology models at scale with a focus on partnerships with large pharma (e.g., Bristol Myers Squibb, Eli Lilly).
- Isomorphic Labs: Isomorphic Labs, an Alphabet/DeepMind spin-out, builds AI foundation models for biology (notably AlphaFold) and pursues partnerships with Novartis and Eli Lilly for AI-driven drug discovery. It is the most direct competitor to Valence's Virtual Cells vision and has substantially larger capital backing.
- Generate Biomedicines: Generate Biomedicines develops generative AI models for protein design and drug discovery. Like Valence, it focuses on generative and predictive biology foundation models and pursues therapeutic programs through both internal pipeline and pharma partnerships (e.g., Amgen, Novartis).
- Exscientia: Exscientia pioneered AI-driven small molecule design and merged with Recursion (Valence's parent) in late 2024, making it both a peer and now a sibling entity within the combined platform. Its Centaur Chemist platform directly competes with Valence's chemistry-relevant Virtual Cells work.
- Schrödinger: Schrödinger provides computational chemistry and physics-based simulation software (including AI/ML-augmented tools) for drug discovery, serving pharma and biotech customers. It is a direct peer in using computational modeling to predict molecular behavior, though it is more software-platform oriented than Valence's biology-first approach.
- Atomwise: Atomwise applies AI (notably its AtomNet platform) to small molecule drug design and has built partnerships with multiple pharma companies. It is comparable to Valence in targeting AI-driven small molecule discovery and predictive molecular interaction modeling.
- XtalPi: XtalPi combines AI, quantum physics, and automated robotics for small molecule and crystal-form drug discovery. It is comparable to Valence in combining physics-based computational chemistry with AI/ML workflows at scale, with significant pharma partnership revenue.
Emerging players
- Owkin: Owkin applies AI/ML to biology and clinical research with a federated learning approach across hospital data. While its primary focus is on clinical/translational AI rather than preclinical drug discovery, it represents a comparable emerging player in AI-driven biology and therapeutic insight generation.
Broad incumbents
- BenevolentAI: BenevolentAI is a publicly listed AI drug discovery company using knowledge graphs and ML to identify novel targets and drug candidates. It represents a broader incumbent in AI-driven target identification and drug discovery with longer commercial history but smaller infrastructure scale than Valence's Recursion-backed setup.
Market position
Strengths5 records
Weaknesses4 records
Competitive moat4 records
Key risks5 records
Key highlights6 records
Customer concentration
Valence Discovery financial estimates
Financial estimateRevenue estimate
Valuation estimate
Valence Discovery leadership team
Management profileNumber of profiles
Profiles3 records
Valence Discovery funding detail
Funding detailFunding overview
Funding rounds5 records
Investors13 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Valence Discovery M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Valence Discovery
What does Valence Discovery do?
Valence Labs develops AI-first computational models—called virtual cells—that predict the functional response of human cells to genetic and chemical perturbations, enabling pharmaceutical and biotech companies to generate, test, and refine therapeutic hypotheses before clinical trials. The platform is built on Recursion's >60 petabytes of interventional biological data and BioHive supercomputer, and includes flagship products such as the TxPert predictive model and the OpenQDC open-source quantum mechanics dataset hub.
When was Valence Discovery founded?
Valence Discovery was founded in 2018. It employs 11 to 50 people.
Where is Valence Discovery based?
Valence Discovery is headquartered in Montréal, Canada, in the North America region.
Who are Valence Discovery's main competitors?
Direct peers on record are Recursion Pharmaceuticals, Insitro, Isomorphic Labs, Generate Biomedicines, Exscientia, Schrödinger, Atomwise and XtalPi. Owkin is listed as an emerging player. BenevolentAI is listed as a broad incumbent.
Does Valence Discovery have an API?
No public API is recorded for Valence Discovery.
What industry is Valence Discovery in?
Valence Discovery's product category is AI Drug Discovery Platform. Its primary akta.pro industry code is HLAGAJAO, AI/ML Platforms for Drug Discovery & Experiment Optimization, with a secondary code of HLAGAJAE, Cheminformatics & Molecular Modeling / CADD. Its NAICS code is 54171 and its SIC code is 8731.