Cradle
Last reviewed
Jun 8, 2026
Sources
9 citations
Review status
Source-backed
Revision
v1 · 1,425 words
Improve this article
Add missing citations, update stale details, or suggest a clearer explanation.
Suggest edit
CtrlK
Last reviewed
Jun 8, 2026
Sources
9 citations
Review status
Source-backed
Revision
v1 · 1,425 words
Add missing citations, update stale details, or suggest a clearer explanation.
Cradle (legally Cradle Bio B.V., styled cradle.bio) is a Dutch-Swiss software company that builds generative AI tools for protein engineering. Its platform lets research-and-development scientists at biotech and pharmaceutical organizations design and optimize proteins, for example to improve a protein's thermal stability, expression, binding affinity, or catalytic activity, by proposing novel amino-acid sequence variants for the customer to validate in the laboratory. Founded in 2021 and headquartered in Amsterdam, the Netherlands, with a second office in Zurich, Switzerland, Cradle sells software rather than developing its own drugs or products, positioning generative AI as general-purpose "software for biology." [1][2][3] The company raised a $73 million Series B round in November 2024 led by IVP, bringing total funding to roughly $100 million. [1][4]
Cradle addresses a core bottleneck in protein design: improving one property of a protein, such as heat resistance or manufacturability, without degrading the others, a search problem made hard by the astronomically large space of possible amino-acid sequences. Conventional protein engineering relies heavily on trial-and-error mutagenesis, with teams developing antibody therapeutics or industrial enzymes typically running dozens of experimental rounds. [2][4]
Cradle's software treats protein sequences like, in CEO Stef van Grieken's framing, "an alien programming language," applying machine-learning models conceptually similar to large language models to suggest sequence variants predicted to exhibit a scientist's chosen properties. Users do not need machine-learning expertise to operate the tool. [2][4] The company markets the platform on efficiency gains rather than novel biology, claiming customer projects progress between roughly 1.2 and 12 times faster and use up to 90 percent fewer experiments than traditional approaches. [1][3] Cradle's applications span medicines, vaccines, diagnostics, enzymes for products such as detergents, sustainable chemicals, food ingredients, and agricultural crop protection. [1][3]
A defining commercial choice is the business model: Cradle is sold as software-as-a-service, with no royalties, revenue sharing, or claims on customer intellectual property. Customers retain all rights to the proteins they engineer and keep control of their data. This distinguishes Cradle from "techbio" companies that use AI to discover and own their own drug pipelines. [1][3][4]
Cradle was founded in 2021, incorporating in late 2021, by a team of five co-founders: Stef van Grieken (CEO), Jelle Prins, Elise de Reus, Eli Bixby, and Harmen van Rossum. [3][5] Van Grieken previously led product work at Google, including at Google Brain and Google X, where his work touched the development of Google's Tensor Processing Unit (TPU) machine-learning hardware; he is also founder and chair of the Open State Foundation and holds an MSc in industrial engineering from the University of Groningen. [6][3] Co-founder Elise de Reus had earlier worked at the synthetic-biology company Zymergen and focused on integrating AI with biological laboratory workflows. [5]
Index Ventures met van Grieken within weeks of the company's incorporation and led a seed round that closed in November 2022. [7] From the outset Cradle operated across two hubs, Amsterdam and Zurich, and built an in-house wet laboratory in Amsterdam, an unusual step for a software company that became central to its strategy. [2][4]
The Cradle platform is built around a generative protein language model trained on a corpus of roughly two billion known protein sequences. [4] A scientist specifies a starting protein and the properties they want to improve, and the model proposes a ranked set of candidate sequence variants to synthesize and test. The customer runs those experiments in their own lab, and the resulting data is fed back to refine the models, so that the system becomes increasingly effective at engineering sequences with the desired functions. This design-test-learn cycle, often described as a "lab-in-the-loop" workflow, is the heart of the product. [1][2][3]
Cradle's own wet lab in Amsterdam plays a distinct role from a customer's lab: rather than running customer projects, it generates large, systematic experimental datasets, effectively A/B testing across many types of proteins, to train and fine-tune the foundational models that underpin every customer's work. Proceeds from the Series B were earmarked partly to expand this lab into additional protein modalities and properties. [1][2] To illustrate model performance, Cradle has reported an internal benchmark on the enzyme T7 RNA polymerase in which about 70 percent of model-generated variants showed increased thermal stability, against success rates typically under 5 percent for conventional methods. [2]
When Cradle closed its Series A in late 2023 it had only a small number of paying customers and projects. By the time of the November 2024 Series B, the company reported around 21 customers running roughly 31 proteins or molecules on the platform, with the count of revenue-generating projects having grown sharply over the prior year. [1][4] Named customers include the pharmaceutical companies Novo Nordisk, Johnson & Johnson (via its Innovative Medicine division), the enzyme and biosolutions firm Novonesis (formerly Novozymes), and the plasma-derived medicines company Grifols. [1][2] Cradle has stated that four of the world's ten largest pharmaceutical companies are among its clients. [1] Reflecting its software-first ambition, van Grieken has framed the long-term goal as putting Cradle's tools into the hands of as many as a million scientists. [4]
Cradle has raised three disclosed venture rounds. Index Ventures led both the seed and Series A, with IVP leading the larger Series B; Kindred Capital has participated across rounds. Reported funding figures and lead investors are summarized below; note that the seed round is reported in euros (about EUR 5.5 million) and is sometimes cited as roughly $5.5 million in US coverage. [3][5][7][8]
| Round | Date | Amount (reported) | Lead investor | Selected other backers |
|---|---|---|---|---|
| Seed | November 2022 | EUR 5.5 million (about $5.5M) | Index Ventures | Kindred Capital |
| Series A | November 2023 | $24 million | Index Ventures | Kindred Capital; angels Chris Gibson and Tom Glocer |
| Series B | November 2024 | $73 million | IVP | Index Ventures, Kindred Capital |
The Series B was announced on November 26, 2024. [1][8] Across these rounds, total capital raised has been reported as roughly $103 million, and commonly described as "over $100 million." [9][1] Coverage of the Series B indicated pricing in the range of several hundred thousand to about $1 million per molecule per year, and Cradle has said it spends on the order of $1 million annually on compute. [4] As of mid-2026, no later round (such as a Series C) had been publicly announced. [9]
Cradle operates in a crowded field of companies applying AI to protein design, but its software-and-services positioning sets it apart from several rivals. Foundation-model builders such as EvolutionaryScale (creator of the ESM family of protein language models) and Basecamp Research are training large protein-generation models and raising substantial venture capital. [9] Vertically integrated drug developers such as Generate Biomedicines and Absci pair AI design with their own laboratories and proprietary pipelines, while Profluent likewise develops AI-designed proteins and therapeutics. Large pharmaceutical companies also run substantial internal protein-engineering efforts. By contrast, Cradle aims to be the tooling layer those existing R&D teams adopt, rather than a competitor for their drug programs. [9][4]
Cradle's emergence reflects a broader shift in which generative AI, having proven useful for text, images, and code, is increasingly applied to biological sequence design, with structure-prediction and sequence-generation advances such as AlphaFold reshaping expectations for what computation can do in the life sciences. The company has been cited as a representative example of European "techbio," including in lists of European startups to watch, and as a case study in selling generative AI as horizontal software for scientific R&D rather than as a vehicle for owning the resulting molecules. [3][9]