Latent Labs
A frontier AI lab building generative foundation models for de novo protein and antibody design.
Website: https://www.latentlabs.com
Cover Block
PUBLIC
| Name | Latent Labs |
| Tagline | A frontier AI lab building generative foundation models for de novo protein and antibody design. [Radical Ventures] |
| Headquarters | London, United Kingdom |
| Founded | 2023 |
| Stage | Seed |
| Business Model | B2B |
| Industry | Deeptech |
| Technology | AI / Machine Learning |
| Geography | Western Europe |
| Growth Profile | Venture Scale |
| Founding Team | Solo Founder |
| Funding Label | $50M+ |
| Total Disclosed | $50,000,000 [TechCrunch, Feb 2025] |
Links
PUBLIC
- Website: https://www.latentlabs.com
- LinkedIn: https://www.linkedin.com/company/latent-labs
- Platform (Beta): https://platform.latentlabs.com/
Executive Summary
PUBLIC
Latent Labs is a frontier AI research lab building generative foundation models for de novo protein and antibody design, a company that merits attention for its technical pedigree and its attempt to shift drug discovery from a wet-lab-heavy process toward a purely computational one [Radical Ventures]. Founded in 2023 by Dr. Simon Kohl, a key architect of DeepMind's Nobel-recognized AlphaFold2 system, the company's core wedge is applying that foundational structural biology expertise to the next step: designing novel proteins from scratch [TechCrunch, Feb 2025]. Its platform, Latent-X, is positioned to allow scientists to computationally generate therapeutic candidates and enzymes, with claimed speed advantages over existing methods like RFdiffusion [decodingbio.substack.com, 2026].
The company has assembled a team of AI researchers and computational biologists from DeepMind, Google, and Microsoft, and has secured significant early capital, raising a total of $50 million across a pre-seed and seed round by early 2025 [TechCrunch, Feb 2025]. Its business model is partnership-driven, aiming to license its technology or co-develop programs with biopharma and industrial biology companies rather than developing its own drug pipeline [Radical Ventures]. Over the next 12-18 months, the key signal to monitor will be the transition from technical validation to commercial validation, specifically the announcement of named, paying partners and the publication of concrete case studies demonstrating the platform's impact on a partner's discovery timeline or pipeline.
Data Accuracy: GREEN -- Core facts confirmed by multiple independent sources including TechCrunch, Radical Ventures, and company materials.
Taxonomy Snapshot
| Axis | Classification |
|---|---|
| Stage | Seed |
| Business Model | B2B |
| Industry / Vertical | Deeptech |
| Technology Type | AI / Machine Learning |
| Geography | Western Europe |
| Growth Profile | Venture Scale |
| Founding Team | Solo Founder |
| Funding | $50M+ (total disclosed ~$50,000,000) |
Company Overview
PUBLIC
Latent Labs was founded in London in 2023 by Dr. Simon Kohl, a former Google DeepMind scientist and a key architect of the AlphaFold2 protein structure prediction system [TechCrunch, Feb 2025] [Radical Ventures]. The company emerged from stealth in February 2025, announcing a total of $50 million in funding across two rounds [TechCrunch, Feb 2025]. Its founding narrative is rooted in a vision to extend the principles of computational protein prediction, as pioneered by AlphaFold, into the generative domain of de novo design, aiming to make biology programmable [Radical Ventures].
Key milestones follow a rapid trajectory from formation to a significant seed round. The company secured an initial $10 million pre-seed round in 2024, followed by a $40 million seed round in early 2025 [TechCrunch, Feb 2025]. By early 2026, the company had released its core generative platform, Latent-X, and announced subsequent model iterations, Latent-X2 and Latent-Y, indicating a continued focus on platform development and speed benchmarks [justainews.com, 2026] [Latent Labs].
Data Accuracy: GREEN -- Confirmed by TechCrunch, Radical Ventures, and company press release.
Product and Technology
MIXED Latent Labs sells computational speed. Its core product, the Latent-X platform, is a generative AI system for designing novel proteins and antibodies directly from a computer browser or API, positioning it as a tool to compress the early stages of drug and enzyme discovery [Radical Ventures]. The company's public claims center on performance benchmarks, specifically a 10x speedup over the established RFdiffusion model for designing a single protein binder and a 56-fold acceleration for a single researcher running design campaigns with its Latent-Y workflow [Latent Labs] [decodingbio.substack.com, 2026].
- Platform architecture. The offering appears as a cloud-based software suite. Customers can access a web interface for interactive design or integrate via an API for automated workflows [Latent Labs].
- Model evolution. The company has iterated its core model, moving from the broadly adopted Latent-X1 to Latent-X2, which is promoted for zero-shot antibody design [justainews.com, 2026] [sofinnovapartners.com].
- Tech stack (inferred from job postings). Open roles for "Software Engineer" and "Member of Technical Staff - Reasoning Workflows" suggest a backend built on modern cloud infrastructure (AWS, GCP, Azure) and a focus on orchestrating complex, multi-step AI reasoning pipelines [Latent Labs].
The technology's stated ambition is to enable "purely computational drug discovery," generating high-confidence therapeutic candidates with minimal initial wet-lab iteration [Radical Ventures]. While the platform is described as being adopted by industry and academic groups, specific customer names and deal structures remain [PRIVATE] [finance.yahoo.com, 2026].
Data Accuracy: YELLOW -- Core product claims are sourced from the company and investor materials; performance benchmarks are cited in third-party technical analysis. Customer adoption and detailed technical specifications are not independently verified.
Market Research
PUBLIC The surge in computational biology funding over the past three years reflects a broader conviction that generative AI can compress the decade-long, billion-dollar timelines of traditional drug discovery.
A precise, third-party TAM analysis for de novo protein design platforms is not yet publicly available. The market is typically sized by analogy to broader biotech AI and drug discovery segments. For context, the global AI in drug discovery market was valued at $1.2 billion in 2023 and is projected to grow at a compound annual rate of 29.4% to reach $5.7 billion by 2028, according to a MarketsandMarkets report [MarketsandMarkets, 2023]. The adjacent market for AI in biotechnology, which includes broader applications in agriculture and industrial biology, is forecast to reach $10.8 billion by 2032 [Precedence Research, 2023]. These figures serve as a proxy for the expansive, but fragmented, opportunity Latent Labs is addressing.
Demand is driven by persistent economic pressures in biopharma. The average cost to develop a new drug now exceeds $2.3 billion, while late-stage clinical failure rates remain high, creating a powerful incentive for tools that can improve early-stage candidate quality and reduce costly wet-lab cycles [Tufts Center for the Study of Drug Development, 2022]. Concurrently, the validation of deep learning for protein structure prediction, exemplified by AlphaFold2, has unlocked investor and corporate appetite for the next wave of generative models that can design, not just predict, biological molecules [Nature, 2021].
Latent Labs operates at the intersection of two high-value application markets: therapeutic protein design and industrial enzyme engineering. The former targets the $280 billion biologics market, where monoclonal antibodies and other engineered proteins are dominant [Evaluate Pharma, 2023]. The latter addresses the sustainable chemicals and materials sector, where engineered enzymes are used in manufacturing processes for biofuels, plastics, and textiles, a market estimated at over $10 billion [Grand View Research, 2023]. This dual focus could provide revenue diversification but also requires navigating distinct customer workflows and success metrics.
Regulatory and macro forces present a mixed picture. Regulatory agencies like the FDA and EMA have established pathways for AI/ML-based software as a medical device (SaMD), but the acceptance of computationally generated novel protein sequences without extensive prior physical data remains an evolving frontier [FDA, 2023]. Geopolitically, initiatives like the U.S. CHIPS and Science Act and the European Innovation Council fund strategic technologies, including bio-AI, which could provide non-dilutive capital and partnership avenues [White House, 2022]. Intellectual property law around AI-generated inventions, however, is unsettled and could impact licensing models.
Therapeutic Biologics Market | 280 | $B
AI in Drug Discovery Market (2028 Projection) | 5.7 | $B
Industrial Enzymes Market | 10 | $B
The chart illustrates the substantial addressable markets adjacent to Latent Labs's core technology. The therapeutic biologics segment represents the primary monetization endpoint, while the industrial enzymes market offers a parallel, non-therapeutic path that may have shorter development cycles and different risk profiles.
Data Accuracy: YELLOW -- Market sizing figures are drawn from established third-party reports for analogous sectors, but a dedicated TAM for de novo protein design platforms is not publicly cited.
Competitive Landscape
MIXED Latent Labs enters a field where computational biology startups are defined by their choice of target, model architecture, and go-to-market, with the company's focus on foundational models for programmable biology placing it against both specialized drug discovery platforms and broader AI research labs.
| Company | Positioning | Stage / Funding | Notable Differentiator | Source |
|---|---|---|---|---|
| Latent Labs | Frontier AI lab building generative foundation models for de novo protein/antibody design. | Seed ($50M total disclosed) | Founder pedigree from AlphaFold2; focus on foundational models for both health & sustainability. | [Radical Ventures], [TechCrunch, Feb 2025] |
| Exscientia | AI-driven drug discovery company with a pipeline of oncology and immunology candidates. | Public (Nasdaq: EXAI) | Integrated platform from AI design to clinical development; multiple programs in Phase 1/2 trials. | Company filings |
| Generate:Biomedicines | Generative biology platform creating novel protein therapeutics. | Venture (Series C $273M in 2021) | Machine learning for generating and optimizing therapeutic proteins without a fixed template. | Company website |
| Cradle Bio | AI platform for protein engineering and optimization. | Seed ($20M in 2022) | Focus on optimizing existing proteins for industrial applications (e.g., enzymes, biomaterials). | [Crunchbase] |
The competitive map breaks into three distinct segments. First, integrated drug discovery platforms like Exscientia and Atomwise have established pipelines and partnerships, competing on the promise of moving candidates into the clinic. Second, generative biology specialists like Generate:Biomedicines and Cradle Bio focus on protein generation but often with a narrower therapeutic or industrial application scope. Third, adjacent substitutes include large tech labs (e.g., Google's Isomorphic Labs) and academic consortia, which compete for talent and set the pace for foundational model breakthroughs but typically lack a dedicated commercial partnership engine.
Latent Labs's defensible edge today rests on two pillars. The first is founder pedigree; Simon Kohl's role as a "key architect" of AlphaFold2 [Radical Ventures] attracts top-tier AI research talent and grants immediate credibility with potential pharma partners. The second is the company's stated ambition to build foundational models for "programmable biology" across both health and sustainability [Radical Ventures, CB Insights], a broader mandate than many therapeutic-focused peers. This edge is perishable, however. The talent advantage decays as other well-funded labs recruit from the same pool, and the broad focus could become a liability if it slows product-market fit in the nearer-term, more lucrative drug discovery vertical.
The company's most significant exposure is to competitors with deeper wet-lab integration and validated pipelines. Exscientia, for example, has moved multiple internally-designed candidates into clinical trials, demonstrating a full-stack capability that Latent Labs has not yet shown. Furthermore, Latent Labs's partnership-led, non-therapeutic development model [TechCrunch, Feb 2025] leaves it vulnerable if potential partners choose to build similar capabilities in-house or partner with a more specialized platform that offers clearer near-term ROI on a specific therapeutic area.
The most plausible 18-month scenario sees the competitive field bifurcating. Winners will be those that convert AI hype into tangible, lab-validated protein designs under partnership deals with disclosed financial terms. A named winner in this scenario could be Generate:Biomedicines if it announces a major partnership for a novel therapeutic modality generated by its platform. Losers will be labs that remain in a perpetual research mode, publishing benchmarks but failing to secure anchor commercial partnerships. Latent Labs would fall into this latter category if, by late 2027, it has not named a flagship biopharma partner or disclosed a lead program advancing toward preclinical studies.
Data Accuracy: YELLOW -- Competitor profiles compiled from public company sources; Latent Labs's differentiation claims are from investor and media reports, not yet from commercial track record.
Opportunity
PUBLIC The prize for Latent Labs is the potential to become the foundational software layer for a new, computationally-driven era of drug discovery and industrial biology, a role that could command platform-level economics across a multi-hundred-billion-dollar value chain.
The headline opportunity is to become the category-defining AI platform for de novo protein design, analogous to what AlphaFold became for structure prediction. Rather than developing a single drug, the company aims to provide the generative models and interfaces that enable partners to design antibodies, enzymes, and therapeutic molecules directly on a computer. The cited evidence points to a reachable, not merely aspirational, outcome. The founder's pedigree as a key architect of AlphaFold2 provides a direct line to the core technology and credibility that underpins modern computational biology [Radical Ventures]. The company's stated focus on building "foundation models" and a platform accessible via web browser and API suggests an infrastructure play from the outset [Latent Labs]. Early performance claims, such as a 10x speedup over RFdiffusion for binder design, indicate a technical wedge that could justify adoption even in a crowded field [decodingbio.substack.com, 2026].
Growth from a promising research lab to a dominant platform could follow several concrete paths, each with identifiable catalysts.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Therapeutics Platform Leader | Latent-X becomes the preferred computational engine for large biopharma partners, moving from pilot projects to multi-program licensing deals. | A major, publicly disclosed partnership with a top-20 pharmaceutical company for a high-value therapeutic area. | The company's explicit go-to-market strategy is to partner with biopharma and biotech firms [TechCrunch, Feb 2025]. Its models are already described as being adopted by industry groups worldwide [finance.yahoo.com, 2026]. |
| Sustainability & Industrial Biology Vertical | The platform gains traction for designing enzymes and proteins for non-therapeutic applications like carbon capture, materials, and agriculture. | A landmark deal with a major chemical or industrial conglomerate, validating the health-and-sustainability focus. | CB Insights characterizes the company as focused on both health and sustainability, providing models for controlling molecular biology processes [CB Insights]. |
| Acquisition by a Tech-Pharma Hybrid | The company's foundational AI talent and IP are acquired by a large-cap firm (e.g., a tech giant with a life sciences division or a pharma company building an internal AI capability). | A strategic investor, potentially already on the cap table, initiates a deeper integration or acquisition discussion. | The investor list includes both deep-tech VCs like Radical Ventures and 8VC, and biopharma-focused firms like Sofinnova Partners, aligning with both tech and biotech exit paths. |
What compounding looks like in this model is a classic data and distribution flywheel. Each successful protein design campaign run by a partner generates valuable, proprietary wet-lab validation data. That data can be used to further refine and improve the foundational models, increasing their accuracy and value for the next campaign. As model performance improves, more partners are attracted to the platform, generating more diverse design problems and validation data. Early hints of this dynamic are present in the company's claims that Latent-Y allows a single researcher to run design campaigns 56-fold faster than expert estimates, suggesting an efficiency gain that could accelerate the feedback loop [Latent Labs]. The more partners that use the platform, the stronger the proprietary dataset becomes, creating a potential data moat that pure open-source models cannot easily replicate.
The size of the win, should the company execute on the "Therapeutics Platform Leader" scenario, can be framed by looking at comparable platform providers in adjacent spaces. For instance, publicly traded companies like Schrödinger (market cap approximately $2.5 billion as of early 2026) provide software and services for computational drug discovery, though not focused on generative de novo design. A more direct, though private, comparable could be Generate:Biomedicines, which raised $273 million in a Series C round at a valuation reported to be over $1 billion [various, 2023]. If Latent Labs successfully transitions from a research lab to a scaled platform with multiple high-value partnerships, a valuation in the low-to-mid single-digit billions is a plausible outcome (scenario, not a forecast). This would represent a significant multiple on the $50 million in seed capital raised, positioning the company as a major standalone asset or a highly strategic acquisition target.
Data Accuracy: YELLOW -- Growth scenarios are extrapolated from stated strategy and early adoption claims; specific catalyst events and comparable valuations are not yet confirmed by the company.
Sources
PUBLIC
[Radical Ventures] Latent Labs Portfolio | https://radical.vc/portfolio/latent-labs/
[TechCrunch, Feb 2025] Founded by DeepMind alumnus, Latent Labs launches with $50M to make biology programmable | https://techcrunch.com/2025/02/12/founded-by-deepmind-alumnus-latent-labs-launches-with-50m-to-make-biology-programmable/
[decodingbio.substack.com, 2026] Latent-X boasts a 10x speedup over RFdiffusion for design of a single binder | https://decodingbio.substack.com/p/latent-x-boasts-a-10x-speedup-over
[Latent Labs] Latent-Y - Latent Labs | https://www.latentlabs.com/latent-y/
[justainews.com, 2026] Latent Labs released Latent-X2 for zero-shot antibody design | https://justainews.com/latent-labs-released-latent-x2-for-zero-shot-antibody-design/
[sofinnovapartners.com] Latent-X1 has been adopted by industry and academic groups worldwide | https://www.sofinnovapartners.com/portfolio/latent-labs/
[finance.yahoo.com, 2026] Latent-X1 has been adopted by industry and academic groups worldwide | https://finance.yahoo.com/news/latent-x1-adopted-industry-academic-120000123.html
[CB Insights] Latent Labs - CB Insights Company Profile | https://www.cbinsights.com/company/latent-labs-technologies
[MarketsandMarkets, 2023] Artificial Intelligence in Drug Discovery Market | https://www.marketsandmarkets.com/Market-Reports/ai-in-drug-discovery-market-151193446.html
[Precedence Research, 2023] Artificial Intelligence in Biotechnology Market | https://www.precedenceresearch.com/artificial-intelligence-in-biotechnology-market
[Tufts Center for the Study of Drug Development, 2022] Cost to Develop a New Drug | https://csdd.tufts.edu/cost-to-develop-a-new-drug/
[Nature, 2021] Highly accurate protein structure prediction with AlphaFold | https://www.nature.com/articles/s41586-021-03819-2
[Evaluate Pharma, 2023] World Preview 2023, Outlook to 2029 | https://www.evaluate.com/thought-leadership/pharma/evaluate-pharma-world-preview-2023-outlook-2029
[Grand View Research, 2023] Industrial Enzymes Market Size, Share & Trends Analysis Report | https://www.grandviewresearch.com/industry-analysis/industrial-enzymes-market
[FDA, 2023] Artificial Intelligence and Machine Learning in Software as a Medical Device | https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-software-medical-device
[White House, 2022] FACT SHEET: CHIPS and Science Act Will Lower Costs, Create Jobs, Strengthen Supply Chains, and Counter China | https://www.whitehouse.gov/briefing-room/statements-releases/2022/08/09/fact-sheet-chips-and-science-act-will-lower-costs-create-jobs-strengthen-supply-chains-and-counter-china/
[Crunchbase] Cradle Bio - Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/cradle-bio
Articles about Latent Labs
- Latent Labs's $50 Million Seed Funds a Browser for Biology — The AlphaFold alumnus is building a web platform to design proteins from scratch, betting on a future where drugs are generated, not discovered.