← News

Basecamp Research raises $140M Series C for AI-designed therapeutics

Basecamp Research said it raised an oversubscribed $140 million Series C led by S32, with NVIDIA and Anthropic’s Anthology Fund among participants, to train next-gen EDEN biological AI models and advance AI-designed therapeutics toward the clinic.

HEALTH desk — a nine-figure check puts Anthropic and NVIDIA money behind AI that designs DNA-scale medicines, not just chatbots — Basecamp is betting you can reprogram a patient’s cells inside the body instead of manufacturing cell therapies one patient at a time.

What the money is for, in the release’s words. The financing is to train a new generation of EDEN models and to move a pipeline of AI-designed therapeutics toward clinical development. Clinical development is the step where a medicine is tested in people, under a regulator’s rules. The release does not name a trial, a hospital, or a first patient. Toward clinical development is a plan. It is not a trial this desk can point to. The page’s own bullets say the money will advance that pipeline, starting with in vivo cell therapy, and expand pharmaceutical partnerships. In vivo means inside a living body. A cell therapy changes a patient’s cells so they do a medical job, often against cancer. Doing that inside the body is the bet. Taking cells out, changing them in a factory, and putting them back is the older way.

Who is on the round. S32 led it. The release names the participants, in this order: Anthropic’s Anthology Fund, Catalio Capital Management, European Tech Collective, Firebrand River Capital, Inception Fund, King Philanthropies, NATO Innovation Fund, NVIDIA, PostScriptum, Redalpine, The Rockefeller Foundation, Singular, Sovereign AI, and True Ventures. NVIDIA and Anthropic’s Anthology Fund are participants. They are not the lead. The release also says additional investment comes from senior leaders across pharma, biotech, and global industry, and it names André Hoffmann, Vice-Chairman of Roche, in that group. A named supporter is not a dollar amount. Do not invent how the $140 million splits.

Three people the release puts on the card. Glen Gowers is co-founder and chief executive. He says the future of medicine is reprogramming the body to repair itself, that the company designs the models and the medicines to teach it how, and that the funding brings the technology closer to patients who have few alternatives. That is his sentence. It is not a cure this desk measured. Andy Conrad, general partner at S32 and former chief executive of Google’s Verily, joins the board. Conrad says AI’s most meaningful opening may be human health, and that Basecamp is building a platform where AI, biological data, and scientific insight meet. Richard Pearce, formerly of Biogen, is appointed chief business officer, to expand pharmaceutical partnerships. The release does not print the month of that appointment. Genetic Engineering & Biotechnology News, in Alex Philippidis’s same-day story, says Pearce was appointed in April and had been Biogen’s head of business development, strategy, and portfolio management. File April, and that Biogen title, as GEN’s. The release’s “formerly of Biogen” is the shorter line.

What EDEN is, and whose superlative it is. EDEN is the company’s biological foundation model. A foundation model is a large model trained on a wide set of examples so it can be aimed at more than one job, not a tool built for a single lab test. The release does not spell the letters. GEN says EDEN is short for “environmentally-derived evolutionary network.” File that expansion as GEN’s. At the center of the platform, the release says, is the Trillion Gene Atlas, which the company calls the world’s largest proprietary biological AI training dataset. The about box calls the same atlas the world’s largest proprietary genomic dataset. Proprietary means the company says it owns the set, rather than training only on a public archive. Both “world’s largest” lines are the company’s. This desk did not rank datasets. The atlas, the release says, was built with partners including NVIDIA, Anthropic, PacBio, and Ultima Genomics. PacBio and Ultima Genomics are companies that read DNA. The data, the release says, was collected through access-and-benefit-sharing partnerships in more than 30 countries across all seven continents. Access-and-benefit-sharing means the samples come with a deal to share some of the value with the places they came from. The release says the models recognize patterns across biology they had not seen, and that they aim at a general picture of how DNA works across life, rather than one scientific task. It says they can propose therapeutic candidates from information about a disease, an approach it says was already shown in a collaboration with Anthropic on Claude Science. The capabilities it names are designing cell and gene therapies, enzymes, and peptides. An enzyme is a protein that does a chemical job. A peptide is a short chain of amino acids, the building blocks of proteins. These lines are the company’s. This desk did not run the model.

The near-term medicine, still the release. Basecamp is applying EDEN to in vivo cell therapy: reprogramming a patient’s cells inside the body. The release says today’s cell therapies are limited in how complex a change they can deliver, and by a manufacturing process that costs hundreds of thousands of dollars per patient. Hundreds of thousands is the company’s description of that older factory step. It is not a Basecamp price. The method it names is pairing EDEN’s ability to design long, complex DNA sequences with large serine recombinases that can insert those sequences into the genome. A recombinase is a protein tool that can place a piece of DNA at a chosen spot. The genome is the full set of DNA in a cell. Large serine recombinases are the class the release names. The aim, in the release’s words, is cell therapies that are more sophisticated, more customizable, and simpler to give, with a chance to change treatment for cancer, autoimmune disease, and more. Autoimmune disease is illness in which the immune system attacks the person’s own tissue. The release says the company has shown strong preclinical results across multiple modalities and disease areas. Preclinical means the work is still in the lab or in animals, not in a human trial the page describes. Strong is the company’s adjective. A modality is a kind of medicine, such as a cell therapy, an enzyme, or a peptide. This desk did not see the data.

Where the company sits, and what the homepage adds that the release does not. The release says Basecamp is headquartered in London, with offices and labs in Cambridge, Massachusetts. The dateline pairs London with Boston. Cambridge is the Boston-area lab city. The company homepage this desk read prints two addresses: Unit 315, Clerkenwell Workshops, 27 Clerkenwell Close, London, EC1R 0AT, and LabCentral 238, 238 Main Street, Cambridge, MA 02142. Those streets are the homepage’s. The same page says the dataset holds over 100 billion novel genes from all seven continents, and over 100 trillion tokens, and that this is the largest biological AI training dataset created to date. A token, here, is one piece of sequence the model trains on. Over 100 billion genes and over 100 trillion tokens are the homepage’s scale. They are not in the release. The homepage also says EDEN is a 28 billion parameter model. A parameter is one of the numbers inside the model. 28 billion is the homepage’s. The homepage says two capabilities, antibiotic design and antigen immunogenicity prediction, are available through Claude. Antigen immunogenicity prediction is a score for whether a piece of a protein is likely to wake up the immune system. That Claude pair is the homepage’s. The release’s Claude Science line is the collaboration, not that product list. The homepage says the therapeutic platforms span gene insertion, gene editing, and peptide design, and that those assets are moving through preclinical development. File that list as the homepage’s.

What GEN adds, and only GEN. Alex Philippidis’s 23 Sep 2026 story does not print an hour. It says Gowers co-founded the company with Oliver Vince, and it gives both men a PhD. The release names Gowers and does not name Vince. File Vince as GEN’s. GEN says EDEN is Basecamp’s new class of foundation models, trained on what the company calls the world’s largest and most diverse evolutionary dataset. “Most diverse” is GEN’s wording of the company’s claim. GEN identifies NVIDIA’s check with NVentures, NVIDIA’s venture arm. The release says NVIDIA. File NVentures as GEN’s name for that check. GEN describes Anthropic’s Anthology Fund as a $100 million fund launched in 2024 and financed by Menlo Ventures, and says companies backed by that fund get access to Anthropic products plus $25,000 in model credits. That is GEN’s background on the fund. It is not a term of this $140 million round. Do not add $25,000 to the raise. GEN says Basecamp raised $60 million in a Series B in 2024 and $20 million in a Series A two years earlier. Those are earlier rounds, GEN’s figures. GEN does not print a lifetime total. Do not add $60 million, $20 million, and $140 million and call the sum a figure GEN stated. GEN says S32 is a Palo Alto firm with about $3.5 billion in assets under management after a $517 million sixth flagship fund earlier this year, according to PitchBook. About $3.5 billion and $517 million are GEN’s, via PitchBook. They are not on the release. GEN also calls Hoffmann a board member of Roche subsidiary Genentech and co-chair of the World Economic Forum. Those titles are GEN’s. The release’s title is Vice-Chairman of Roche. GEN’s investor paragraph names Catalio, European Tech Collective, Firebrand River Capital, King Philanthropies, NATO Innovation Fund, Redalpine, The Rockefeller Foundation, Singular, Sovereign AI, and True Ventures, and it does not repeat Inception Fund or PostScriptum. The release’s list is the roster. Do not drop those two names because a later paragraph is shorter.

Plain English for the rest of the card: Series C = a later private round. oversubscribed = more investors wanted in than the company took. $140 million = the raise the release announced. It is not a valuation. in vivo = inside the body. cell therapy = a medicine that changes a patient’s cells. foundation model = a large model trained for many jobs. EDEN = the company’s biological model. GEN says the letters stand for environmentally-derived evolutionary network. Trillion Gene Atlas = the company’s private DNA training set. access-and-benefit-sharing = a deal to share some value with the places the samples came from. recombinase = a protein tool that inserts DNA. genome = the full DNA in a cell. preclinical = lab or animal work, not a human trial the release describes. enzyme = a protein that does a chemical job. peptide = a short protein chain. modality = a kind of medicine. parameter = one number inside a model. The homepage’s 28 billion is that count. token = one piece of sequence the model trains on. The homepage’s “over 100 trillion tokens” is that scale. NVentures, the April date for Pearce, Oliver Vince, the $60 million and $20 million earlier rounds, the PitchBook figures, and the extra Hoffmann titles are GEN’s. This filing is the 23 Sep announcement. It is not a drug approval.

PRIMARY here: Basecamp Research’s 23 Sep 2026 PR Newswire release, stamped 09:00 ET and datelined London and Boston — Tier A PRIMARY, the company’s own record on the wire. The company homepage is product-home context, not a second originating newsroom. GEN’s same-day story by Alex Philippidis is independent coverage, not a substitute primary. The $140 million oversubscribed Series C, the S32 lead, the named participant list, Hoffmann as Vice-Chairman of Roche, Gowers as co-founder and chief executive, Conrad joining the board as S32 general partner and former chief executive of Google’s Verily, Pearce as chief business officer formerly of Biogen, the EDEN training plan, the in vivo cell-therapy focus, the hundreds-of-thousands manufacturing line, large serine recombinases, the cancer and autoimmune aim, the strong-preclinical sentence, the Trillion Gene Atlas and its partners, the 30-plus countries and seven continents, Claude Science, the cell, gene, enzyme, and peptide capabilities, and London plus Cambridge, Massachusetts are the release’s. The street addresses, the over-100-billion-gene line, the over-100-trillion-token line, the 28 billion parameters, and the Claude antibiotic and immunogenicity pair are the homepage’s. The EDEN expansion, Vince, the PhDs, Pearce’s April date and Biogen business-development title, NVentures, the Anthology Fund’s $100 million and $25,000-credit background, the $60 million Series B and $20 million Series A, the PitchBook assets and fund size, and the Genentech and World Economic Forum titles are GEN’s. NOT claimed: a valuation, a per-investor split, a lifetime funding total this desk added up, that preclinical results were audited here, that a human trial has started, that $25,000 in model credits is part of the $140 million, a stock tip, or investment advice. Distinct from the already-filed enveda-311m, arpa-h-circle, and powerful-medical-queen-of-hearts-fda-de-novo.

RELATED

ONLINE

article thread

guidelines

warming…

warming…

On 23 Sep 2026, Basecamp Research said it raised $140 million in an oversubscribed Series C. The record is the company’s release on PR Newswire, “Basecamp Research raises $140M to advance AI-designed therapeutics,” stamped Sep 23, 2026, 09:00 ET. 09:00 Eastern is 1:00 p.m. UTC. The dateline is LONDON and BOSTON, Sept. 23, 2026. The release calls Basecamp a frontier AI company developing AI-designed medicines. A Series C is a later private funding round, after earlier checks. Oversubscribed means the company says more investors wanted in than it took. $140 million is the amount the release says came in. It does not print a valuation, which would be the price this round puts on the whole company, and it does not print how much each investor wrote. These lines are the company’s, on the wire. This desk did not see a term sheet.

Sources