AI mega-rounds are spilling from drug discovery to robots to chips — somebody’s writing very large checks all over the stack. Welcome to Startup Fundraising. Today we’re running through five rounds totaling more than three billion dollars, and asking if the story still holds at those price tags. Bio, robots, inference hardware, self-improving AI, security — it feels like somebody spun a wheel and funded wherever it landed. I want to know what each of these companies actually has to show for it. Isomorphic Labs is the biggest one at two-point-one billion. Let’s start there. This one's from BioSpace:
LONDON, May 12, 2026 /CNW/ -- Isomorphic Labs, an AI-first drug design and development company, today announced it has raised $2.1 Billion in Series B funding. This latest round of investment will accelerate the company's evolution from pioneering novel AI models to applying them at scale.
The financing round is led by Thrive Capital, and includes participation from existing backers Alphabet and GV alongside new investors MGX, Temasek, CapitalG, and the UK Sovereign AI Fund, significantly expanding Isomorphic Labs' global capital base.
Isomorphic Labs — the Alphabet spinout on the AlphaFold lineage — just closed a two-point-one billion dollar Series B, led by Thrive Capital, with Alphabet, GV, CapitalG, Temasek, MGX, and the UK Sovereign AI Fund in the round. BioSpace had the announcement. And worth flagging: Alphabet is both a prior backer and represented here through CapitalG, so Google’s fingerprints show up in more than one place on that cap table. Thrive leading is the real signal — they don’t do biotech for fun. Two-point-one billion Series B. That’s a number that needs a pipeline, not a platform deck. What drugs are actually moving toward the clinic, and when does somebody start asking why we funded a very expensive protein-folding demo? The use-of-funds language is the usual fog: scale the engine, expand the pipeline, hire world-class talent. That’s not a plan, that’s a mood board. I want the therapeutic areas, the phase-one targets, and whether any pharma partner is actually on the hook for milestone payments. Here's Brian Petersen at Aidaily Post:
Recursive has stepped out of stealth with a $650 million war chest and a $4.65 billion valuation, aiming to build AI that improves itself. The round was led by GV and Greycroft, while AMD Ventures and Nvidia also wrote checks.
Out of stealth, six-fifty million, four-point-six-five billion valuation — GV and Greycroft led, and AMD Ventures plus Nvidia wrote checks too. Richard Socher, formerly Salesforce AI chief, is co-founding with Tim Rocktäschel out of Google DeepMind. No product, no published results, and the pitch is that they’re automating the scientific method to reach superintelligence. What has to be true in two years for this to be worth four-point-six-five? Because right now the deliverable is a Stanisław Lem citation. To be fair, the cap table does read like a serious bet. GV doesn’t lead a round this size on vibes alone, and Nvidia writing a check is basically a hardware pre-commitment. The terms we don’t have are the part I’d want to see. Strong pedigree, I’ll give them that. But “automating the scientific method” is the kind of line that sounds magnificent until someone asks what ships first, who it ships to, and when. Here's Greg Bock at The AI Insider:
Industrial robotics startup Mind Robotics has raised $400 million in new funding led by Kleiner Perkins, bringing total investment in the company to more than $1 billion.
According to Mind Robotics, the round included new investors Meritech Capital, Redpoint Ventures, SV Angel, Incharge Capital, A-Star Capital and Garuda Ventures, alongside existing backers Accel, Andreessen Horowitz, Eclipse, Prysm Capital, Bain Capital Ventures and Greenoaks.
Mind Robotics closed a $400 million round led by Kleiner Perkins — Greg Bock had it in The AI Insider yesterday — pushing total raised past a billion. The investor list is long: Meritech, Redpoint, SV Angel as new names, and basically every tier-one on the existing side from a16z to Greenoaks. A billion dollars in, and the flagship customer is Rivian — a company that’s fighting for its life on unit economics. That’s the deployment and training partner they’re leading with? To be fair, dexterous manufacturing automation is a real problem, and Rivian’s assembly lines are a real testbed. But with twelve-plus investors in this round, I want to know who actually has board conviction here versus who just got allocation. Kleiner led, so somebody has conviction. Everyone else on that list — a few of them are just not missing the robotics moment. What I need to know is whether there’s recurring revenue attached to Rivian, or if this is still a glorified pilot. Here's Amit Chowdhry at Pulse 2.0:
Fractile announced it has raised $220 million to accelerate the development of a new generation of inference hardware designed to dramatically increase the speed and efficiency of frontier AI systems.
Founded in 2022, Fractile was created on the belief that the long-term impact of advanced AI systems would ultimately be constrained not by model quality alone, but by the time required to generate useful outputs.
Fractile just closed two-twenty to build inference hardware. Founded in 2022, so it’s only about three years old, and they’re already swinging at frontier-scale AI chip infrastructure. Okay, the inference bottleneck thesis is real — I’ll give them that. But two-hundred-twenty million to build custom silicon from a four-year-old company? The graveyard of AI chip startups is extremely well-populated. The pitch is basically that models are getting better at reasoning by spending more compute at inference time, not just training time — and nobody’s hardware is really built for that workload at scale. That’s a sharp wedge if you can execute. What I want to know is whether there’s a customer commitment behind this round, or if two-twenty is going straight into the fab pipeline on a prayer. This one's from SecurityBrief:
Exaforce has raised USD $125 million in a Series B funding round, bringing the cyber security company's total funding to USD $200 million.
Investors in the round included HarbourVest, Peak XV, Mayfield, Khosla Ventures, Seligman Ventures and AICONIC. The company will use the funds to further develop its platform and expand internationally, including in Japan and Europe.
Exaforce just closed a $125 million Series B — total funding is now $200 million, with a $75 million Series A just a year ago. The syndicate includes Mayfield, Khosla, Peak XV, and HarbourVest. This is a real cap table, not a friends-and-family stack. A hundred and twenty-five million in twelve months after a seventy-five million A — that’s a fast burn clock starting. So what does the money actually do? They’re expanding to Japan and Europe, which is sales infrastructure spend, not R&D. I want to know what the ARR looks like before they’re flying salespeople to Tokyo. The product angle is a real-time security knowledge graph — ingesting identity, permissions, cloud activity, code, all live rather than after the alert. That’s a real architectural bet against the existing SIEM-and-triage model. Whether it holds under enterprise load is the question. Every AI security vendor right now is saying, “we see things before the alert fires.” What makes this one stick is whether a CISO can cut headcount or actually close incidents faster — not a demo graph, real ticket-closure rates. I haven’t seen that number anywhere. If Startup Fundraising helps you stay sharper on the funding landscape, take a moment to subscribe and leave a review wherever you’re listening. It really helps other founders, operators, and investors find the show.
If something from today’s briefing is worth a closer look, we’ve put links to every story in the show notes so you can follow the details for yourself. That’s Startup Fundraising for today. This is a Lantern Podcast.