Chemify Secures £22M to Scale AI-Powered Chemistry and Build Its Next Chemifarm

Chemistry has become another frontier in the effort to connect artificial intelligence with the physical world. Glasgow-based Chemify is now preparing to expand the infrastructure needed to do exactly that, securing £22 million ($29.7 million) in grants to scale a platform that combines AI-driven molecular design with automated robotic synthesis.

The funding includes £16 million from Scottish Enterprise and £6 million from the UK Government and will support the development of Chemify’s second-generation Chemifarm, an automated facility designed to increase the number of chemical experiments the company can design, execute and learn from.

At the same time, Chemify has extended its previously announced Series B financing to £51.9 million ($70 million), bringing the company’s total funding secured to more than £110 million ($150 million). The original Series B was co-led by Wing Venture Capital and Insight Partners, with participation from 8VC and existing investors including Triatomic Capital, Blueyard, Rockspring and Eos.

The expansion is intended to more than double Chemify’s Chemputation capacity and support a broader ambition: turning chemistry into an increasingly programmable process rather than one built around largely manual laboratory workflows.

From AI Predictions to Molecules That Can Actually Be Made

One of the persistent challenges facing AI-driven chemistry is the difference between designing an interesting molecule computationally and demonstrating that it can actually be manufactured.

Generative models can search enormous chemical spaces and propose structures that appear promising based on predicted properties. But a molecule existing inside a model does not necessarily mean chemists have a practical synthesis route for producing it.

Chemify has structured its technology around closing that gap.

Its core approach, known as Chemputation, combines machine learning, chemical software, a universal chemical programming language and robotic laboratory systems. A target molecule can move from computational design through route planning and ultimately into automated physical synthesis.

Chemify’s route optimization system, called ASSEMBLER, searches a database of reaction classes that the company says have already been validated and automated. The resulting synthesis instructions can then be encoded using χDL, Chemify’s hardware-agnostic chemical programming language.

Those instructions are executed by robotic systems that can handle processes including reaction synthesis, workup, purification and analytical assessment.

In effect, Chemify is attempting to create something closer to a software execution environment for chemistry: AI proposes what to make, software determines how it could be made, and automated laboratory infrastructure executes the instructions.

Chemify Genesis Brings the Experimental Loop Back Into AI

The company is also introducing Chemify Genesis, which extends this infrastructure into a closed-loop discovery system.

Genesis is designed around an “ask, make, test” cycle. Researchers define a scientific objective and constraints, the system proposes and synthesizes candidates, experimental systems measure the results, and those results feed back into subsequent rounds.

That feedback is important because chemical experiments produce useful information even when they fail.

Rather than storing only successful reactions, Chemify says Genesis records successes, partial outcomes and unsuccessful experiments alongside information such as procedures, conditions and analytical data. The resulting dataset can gradually define not only what chemistry works, but where the boundaries of practical synthesis lie.

This forms the basis of what Chemify calls a world model for chemistry.

The terminology has become increasingly common across AI research, where world models attempt to build representations of how an environment behaves so an AI system can make better predictions about future actions. In Chemify’s case, the environment is chemical space.

Instead of learning exclusively from publications or historical datasets, Genesis can incorporate observations generated directly through physical experiments.

That creates a potentially valuable feedback loop: every reaction generates additional evidence that can influence which molecules or synthesis routes the system attempts next.

Building a Larger Chemifarm

Scaling that loop requires substantially more physical infrastructure than deploying another software model in the cloud.

Chemify opened its first Chemifarm in Glasgow in 2025 as a facility dedicated to automated molecular design and synthesis. The company’s next-generation facility is expected to come online later in 2026 with a larger collection of automated systems, higher throughput and expanded synthesis capabilities.

This physical requirement explains Chemify’s use of the term “Chemistry Hyperscaler.”

The comparison is loosely modeled on cloud computing hyperscalers, where enormous concentrations of standardized computing infrastructure make it possible to allocate computing capacity on demand. Chemify is effectively proposing an analogous model for automated chemistry, where organizations reserve access to standardized robotic synthesis capacity rather than building every laboratory workflow internally.

Genesis is designed to provide partners with access to both the software layer and physical Chemifarm capacity. Chemify says individual partner models and experimental data are maintained within separate protected domains rather than pooled across customers.

The company has also indicated that the Glasgow expansion is unlikely to be its last. Additional Chemifarm facilities are planned internationally, including in the United States.

£22M Grant Supports a Much Larger Glasgow Expansion

The public funding represents one component of a significantly larger investment in Chemify’s Scottish operations.

Scottish Enterprise says the £22 million grant package accompanies £67.9 million being contributed by Chemify, bringing the overall expansion project to £89.9 million.

The company also plans to substantially increase its workforce.

Chemify currently employs approximately 152 people in Scotland and expects that figure to grow to around 450 over the next three years. The expansion is expected to create as many as 300 jobs while safeguarding another 100.

Its new global headquarters and research and development center will be located at the Health Innovation Hub in Govan, within the Glasgow Riverside Innovation District near Queen Elizabeth University Hospital.

The expansion further strengthens Glasgow’s role as the center of Chemify’s operations even as the company begins preparing for international deployment of its technology.

A Different Scaling Problem for AI

Much of the AI infrastructure race has focused on GPUs, data centers and access to electricity. Companies such as Chemify illustrate another emerging infrastructure problem: AI systems operating in scientific fields eventually need to interact with physical experiments.

For chemistry, generating another million virtual molecules is comparatively easy. Manufacturing those compounds, determining whether they actually exhibit the predicted properties and feeding the experimental results back into a model is substantially harder.

Chemify’s bet is that automating this physical layer will make AI-driven chemical discovery more useful.

Its model also changes what “scale” means in computational chemistry. More compute remains valuable, but discovery capacity increasingly depends on the number of experiments that can be reliably executed, analyzed and converted back into machine-readable data.

The second Chemifarm and Genesis are designed around precisely that constraint.

If Chemify can make automated chemical experiments increasingly repeatable and programmable, the company could give AI systems something that purely computational discovery platforms cannot provide on their own: a continuous connection between digital predictions and physical chemical reality.

With £22 million in new public funding, an expanded £51.9 million Series B and more than £110 million secured overall, Chemify now has considerably more capital to test whether that model can move from an ambitious approach to digital chemistry into infrastructure capable of operating at industrial scale.