Basecamp Research raises $140m Series C with Nvidia and Anthropic backing
S32 leads the round for the London company that builds its own biodiversity sequencing dataset instead of training on public sequence databases.
Basecamp Research, a London biological-AI company, announced on 23 September 2026 an oversubscribed $140 million Series C led by S32, the Palo Alto firm founded by Bill Maris. Reuters put the post-money valuation at $800 million. NVentures, Nvidia's venture arm, and the Anthology Fund, run by Anthropic with Menlo Ventures, came in as new backers, alongside the NATO Innovation Fund, Redalpine, Catalio Capital Management, the Rockefeller Foundation, European Tech Collective and King Philanthropies. True Ventures and Singular returned. André Hoffmann, vice chairman of Roche, invested as an individual.
The round takes total capital raised to roughly $220 million, following a $60 million Series B in October 2024. The company was founded by Glen Gowers, chief executive, and Oliver Vince, and runs out of Clerkenwell in London with a second office in Cambridge, Massachusetts. Its premise is that public sequence repositories are too narrow and too duplicated to train useful biological models, so it collects its own genomic material from environments around the world.
Two assets carry the thesis. EDEN is the company's biological foundation model. The Trillion Gene Atlas is the dataset behind it, assembled with Nvidia, Anthropic, PacBio and Ultima Genomics. Tech Funding News reports that the sampling runs through access and benefit-sharing agreements in more than 30 countries, the legal framework that governs who may take genetic material out of a country and on what terms, and that the company has six disclosed programmes, all preclinical.
The money goes towards moving AI-designed candidates, including peptides, enzymes and in vivo cell therapies, closer to the clinic. “We believe the future of medicine lies in reprogramming the body to repair itself. We design the models and the medicines to teach it how,” said Gowers. No Basecamp programme has yet entered clinical trials.
Why it matters for AgTech founders: the defensible part of this company is a physical logistics operation, not a model. Anyone can fine-tune on public sequence data, and that is exactly why public sequence data is worth little. Basecamp spent years running field expeditions and negotiating country-level benefit-sharing agreements, and it is that paperwork and those freezers that investors just valued at $800 million. Ag biologicals sit on the same ground. Strain discovery, enzyme prospecting and microbiome work all depend on sampling rights that take years to secure and cannot be replicated by a competitor with a better GPU budget. If your discovery platform's only input is a licensed database, you have a workflow, not a moat. The question worth putting to your own roadmap is which part of your data supply is contractually yours and how long a well-funded rival would need to rebuild it.
Basecamp Research, a London biological-AI company, announced on 23 September 2026 an oversubscribed $140 million Series C led by S32, the Palo Alto firm founded by Bill Maris. Reuters put the post-money valuation at $800 million. NVentures, Nvidia's venture arm, and the Anthology Fund, run by Anthropic with Menlo Ventures, came in as new backers, alongside the NATO Innovation Fund, Redalpine, Catalio Capital Management, the Rockefeller Foundation, European Tech Collective and King Philanthropies. True Ventures and Singular returned. André Hoffmann, vice chairman of Roche, invested as an individual.
The round takes total capital raised to roughly $220 million, following a $60 million Series B in October 2024. The company was founded by Glen Gowers, chief executive, and Oliver Vince, and runs out of Clerkenwell in London with a second office in Cambridge, Massachusetts. Its premise is that public sequence repositories are too narrow and too duplicated to train useful biological models, so it collects its own genomic material from environments around the world.
Two assets carry the thesis. EDEN is the company's biological foundation model. The Trillion Gene Atlas is the dataset behind it, assembled with Nvidia, Anthropic, PacBio and Ultima Genomics. Tech Funding News reports that the sampling runs through access and benefit-sharing agreements in more than 30 countries, the legal framework that governs who may take genetic material out of a country and on what terms, and that the company has six disclosed programmes, all preclinical.
The money goes towards moving AI-designed candidates, including peptides, enzymes and in vivo cell therapies, closer to the clinic. “We believe the future of medicine lies in reprogramming the body to repair itself. We design the models and the medicines to teach it how,” said Gowers. No Basecamp programme has yet entered clinical trials.
Why it matters for AgTech founders: the defensible part of this company is a physical logistics operation, not a model. Anyone can fine-tune on public sequence data, and that is exactly why public sequence data is worth little. Basecamp spent years running field expeditions and negotiating country-level benefit-sharing agreements, and it is that paperwork and those freezers that investors just valued at $800 million. Ag biologicals sit on the same ground. Strain discovery, enzyme prospecting and microbiome work all depend on sampling rights that take years to secure and cannot be replicated by a competitor with a better GPU budget. If your discovery platform's only input is a licensed database, you have a workflow, not a moat. The question worth putting to your own roadmap is which part of your data supply is contractually yours and how long a well-funded rival would need to rebuild it.
