Biomatter announced on November 12, 2026 that it had closed an $18 million Series A to expand its AI enzyme design business. The Vilnius, Lithuania company plans a larger design-and-testing facility and a stronger US presence. For researchers following AI in biology, the investment connects computational design with expanded capacity to build and test enzymes in the laboratory.
Astanor leads the financing
The round was oversubscribed and led by Astanor, with existing investors UVC Partners, Inventure and Metaplanet participating, according to Biomatter’s November 12 announcement. It describes a completed financing, with proceeds allocated to laboratory infrastructure, commercial expansion, the computational platform and the company’s enzyme dataset.
COBALT’s transaction notice, also dated November 12, confirms that it advised Astanor as lead investor. The law firm identifies the same participating investors and calls this Biomatter’s largest financing to date. Its Lithuania advisory team was led by partner Akvilė Bosaitė and associate partner Aurelija Balčiūnė.
TechFundingNews reported that Biomatter did not disclose its valuation or total funding. The $18 million figure therefore describes this Series A round, rather than a valuation or a cumulative fundraising total.
A new Vilnius lab is planned for 2027
The infrastructure plan centres on a 2,000-square-metre enzyme design and testing facility in Vilnius. COBALT says the facility is scheduled to open in 2027 and will bring more laboratory development and testing in-house. It also confirms plans to expand Biomatter’s commercial team and US presence, where approximately half its customers are based.
| Development | Announced scope | Status |
|---|---|---|
| Series A | $18 million, led by Astanor | Closing announced November 12, 2026 |
| Vilnius facility | 2,000 square metres for design and testing | Opening planned in 2027 |
| Development cycles | Move from weeks to days | Company objective for the expanded operation |
| Commercial expansion | Larger team and stronger US presence | Planned use of proceeds |
These milestones are described in the adviser’s account of the transaction. The facility opening remains a future step, and the shorter development cycle is an expected outcome of bringing more work in-house.
Enzyme design combines computation with experiments
Biomatter describes its approach as starting with the chemical reaction a customer needs, rather than taking a natural enzyme as the starting point. Its announcement says each design is tested in the laboratory before delivery. The company reports more than 50 custom enzymes delivered to more than 25 enterprise customers across Europe, the United States and Japan.
In its November 12 report, Tech.eu explains that Biomatter combines computational models and laboratory testing around customer requirements. Experimental results expand a proprietary dataset used in subsequent development programmes. Biomatter says its current programmes take weeks; the move to days is the ambition associated with the new facility.
For researchers in protein engineering, this makes the relationship between design and experimental characterisation central to the story. The funded expansion includes the physical work needed to evaluate candidate enzymes, alongside further development of the computational platform.
Karpus links the expansion to earlier research
In a November 12 interview with Tech Philomaths, CEO and co-founder Laurynas Karpus recalled work in 2019 demonstrating algorithms that generated novel, functional enzymes, subsequently published in Nature Machine Intelligence in 2021. He presented newer general AI advances as a way to get more from Biomatter’s own enzyme models and laboratory data.
Karpus distinguished those proprietary resources from general large models available to other developers. He identified a new enzyme research and development centre in Vilnius, team growth and US expansion as priorities for the Series A. He also said the preceding funding period had demonstrated the usefulness of the company’s computational and physical platform to industrial customers.
That account places the financing after earlier research and customer development. It also highlights a practical issue for academic AI research: access to a general model and access to the experimental data behind a specialised system are separate matters. CASRAI’s definition of training data provenance explains the associated question of where model inputs originate.
Commercial adoption is the investors’ stated rationale
TechFundingNews identifies customers including AstraZeneca, BASF, Kirin Holdings, Neogen, Sanofi and Tereos. It reports that some relationships have expanded from individual projects into programmes involving multiple enzymes over several years. These customer and delivery figures are company-reported measures of commercial activity.
The outlet also reports the investors’ rationale: UVC Partners partner Oliver Schoppe pointed to enzymes reaching customers’ commercial products, while Astanor partner Harry Briggs emphasised proprietary experimental data and repeat customers. Those are the backers’ assessments of the business, rather than independent measurements of enzyme performance.
What the investment means for research partnerships
For universities and research offices, the concrete development is additional financing for a commercial partner combining enzyme design, proprietary data and laboratory delivery. The reported multi-year customer programmes make the scope of a collaboration relevant alongside the properties of any individual enzyme. A sponsored research agreement provides background on how institutions define research work with an external sponsor; the financing announcement itself sets out company expansion plans.
The next dated milestone is the planned 2027 Vilnius facility opening. As Tech.eu reports, the company intends that facility to shorten the design-and-test cycle while continuing investment in its platform and dataset. Whether the expanded operation achieves development cycles measured in days remains a future outcome.
Sources
- Biomatter, company-issued EQS announcement via Webdisclosure, November 12, 2026. Checked 2026-11-16.
- COBALT, Astanor transaction advisory notice, November 12, 2026. Checked 2026-11-16.
- Tech.eu, Tamara Djurickovic, Biomatter lands $18M to speed up enzyme development with AI, November 12, 2026. Checked 2026-11-16.
- TechFundingNews, Abhinaya Prabhu, Biomatter lands $18M from Astanor and UVC Partners to build a Vilnius enzyme lab, November 12, 2026. Checked 2026-11-16.
- Tech Philomaths, interview with Laurynas Karpus, November 12, 2026. Checked 2026-11-16.







