Silicon Valley is going physical, and the money is following.
After years of betting almost everything on software, investors are redirecting billions toward a sector that, until recently, was seen as too expensive, too risky, and too hard: robotics. And we are not talking about a shy or experimental move here. We are seeing a structural shift, the kind of change that happens once in a generation and completely redefines which companies will matter over the next ten, twenty years.
The numbers speak for themselves. Startups in physical AI raised $16.3 billion across 492 deals in the first quarter of 2026 alone, according to PitchBook data. To put that in perspective, that volume already exceeds the total invested in robotics during entire years of the last decade. This is serious money, from serious funds, going to companies that are building things you can touch, move, and program.
It is no coincidence. Three forces are pushing this movement at the same time: hardware costs have dropped significantly in recent cycles, labor shortages keep growing in developed economies, and there is increasing pressure to bring manufacturing back to the United States. When these three variables come together, the result is exactly what we are seeing now: capital flowing at scale toward those building for the physical world.
The outcome of all this is a new generation of companies developing concrete solutions to real problems, and the most attentive investors are already positioned well before most people realize what is happening. To map out this landscape, Business Insider consulted 14 investors from the robotics ecosystem, including names like Bain Capital Ventures, Sequoia Capital, and Bessemer Venture Partners. Each one recommended two startups: one from their own portfolio and another with no financial interest involved. What emerged from this curation is a list that shows the real scale of this boom. 🤖
Why robotics became a priority now
For a long time, robotics was confined to highly controlled industrial environments, like automotive assembly lines where every millimeter of the process was predictable and repeatable. The problem is that the real world does not work like that. Outside those controlled bubbles, robots simply could not handle variation, improvisation, or ambiguous contexts. Then artificial intelligence entered the picture, and everything changed. Language models and computer vision systems have advanced so much in recent years that today it is possible to train a robot to understand context, adapt behavior, and make decisions in real time, something that was practically science fiction not long ago.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
This technological leap came at a perfect moment. The cost of sensors, actuators, and embedded processors has dropped consistently. What once required a university research budget now fits into the business plan of an early-stage startup. This opened up space for entrepreneurs with good ideas to actually build functional prototypes, raise their first rounds, and evolve rapidly. The innovation cycle has accelerated in a way that even industry veterans acknowledge as unprecedented. It is worth noting, however, that robotics is still a young field and much of this technology remains unproven at scale. Researchers interviewed by Business Insider went as far as classifying humanoids as a fantasy product with little short-term viability, and they warn that general-purpose domestic robots are still many years away.
But perhaps the most underestimated factor in this equation is the human one. The available workforce for repetitive physical tasks is shrinking in virtually every developed market. Companies that need these roles are not waiting around for a political solution. They are buying robots, and they are willing to pay well for them.
The profile of startups leading this movement
What stands out in the list curated by investors is the diversity of approaches. There is no single company model dominating this moment. There are startups focused on developing what the industry calls a generalist brain for robots, meaning artificial intelligence systems capable of controlling different types of hardware without needing to be reprogrammed from scratch for each new task. Names like FieldAI, Generalist, and Skild AI fit into this category. This approach is especially attractive to investors because it allows a single software solution to be sold to multiple manufacturers, expanding the potential for scale without proportionally increasing costs.
Generalist, for example, was founded by roboticists who came from Google DeepMind and Boston Dynamics. To teach its AI about the physical world, the company created low-cost grippers that mimic robotic hands and distributed thousands of them around the world, collecting over 500,000 hours of training data. Meanwhile, Skild AI, founded by former Carnegie Mellon professors, has accumulated $2.2 billion in funding and is working with Nvidia and Foxconn to automate production lines for Blackwell chips.
There is also a significant group of companies developing humanoids, robots with a roughly human form that can operate in environments designed for people, like warehouses, factories, and even homes. Unitree, based in China, showed that it is possible to build humanoids relatively cheaply and at scale, having shipped over 5,500 units in 2025. The company is preparing for an IPO in Shanghai with a valuation of around $9 billion. Sunday Robotics, on the other hand, bet on the domestic environment with the Memo robot, which reportedly folded laundry in unfamiliar homes with a success rate above 99%.
Beyond the generalists and humanoids, the list also includes companies with a very well-defined vertical focus. Metallurgy, drug development, and precision agriculture stand out as sectors with real and growing demand for intelligent physical automation. Dash Bio, co-founded by a former Moderna executive, automates laboratory testing for pharmaceutical development and claims to deliver results roughly ten times faster than the industry average. In agriculture, companies like Synphony and Upside Robotics are tackling serious harvesting and fertilization challenges. Synphony started by training robots to pick strawberries, while Upside develops small solar-powered robots that apply fertilizer directly at the root of plants, with the potential to reduce usage by up to 70%. These are massive markets, with reasonable margins and clients who already understand the value of what they are buying.
What the investor-selected companies have in common
Looking at the recommendations as a whole, some patterns become pretty evident. The companies that showed up most frequently in investor picks share a characteristic that might seem obvious but is actually hard to pull off in practice: they combine real technical depth with problem clarity. It is not just about having a team of brilliant engineers doing advanced research. It is about having that team working on a specific pain point that real customers are willing to pay to solve right now, not in some hypothetical future.
A good example of this logic is Cobot, led by Brad Porter, who previously ran Amazon Robotics and served as CTO of Scale AI. Its Proxie robot has already logged nearly 13,000 hours of operation and moved over 154,000 carts for clients like Mayo Clinic and Maersk. Another interesting case is Gecko Robotics, out of Pittsburgh, which builds climbing robots to inspect power plants, pipelines, and military vessels, and in March closed a contract worth up to $71 million with the U.S. Navy.
Another pattern that stands out is the stage of the selected companies. The list is quite diverse on this front, ranging from companies that recently came out of Y Combinator, still in the early stages of their go-to-market journey, to businesses that already have recurring revenue and are gearing up for an IPO. This is an important signal: the robotics boom is not a speculative bet on technologies that will only exist ten years from now. It is a market generating real business across multiple stages of maturity at the same time, which reduces risk for the ecosystem as a whole.
It is also worth highlighting the growing importance of the training data question, which has become one of the biggest bottlenecks in robotics. Companies like XDOF, Mecka AI, and Avatar Robotics were born precisely to solve this, collecting real-world information about how humans interact with objects and environments. XDOF, pointed out by three different investors, understands that the collection hardware itself, from cameras to tracking algorithms, directly affects the final quality of the data. As Amanda Huang from Bain Capital Ventures put it: garbage data in, garbage results out.
The curation done by Business Insider also serves as an interesting barometer for how investors themselves view the sector. The fact that each participant recommended one portfolio company and one company with no financial ties shows that the enthusiasm is not purely corporate. When a partner at Sequoia or Bessemer points to a company that will not generate a direct return for their fund, that says a lot about the conviction that exists around this market.
The geopolitical question weighs in too
A theme that comes up strongly among several of the recommended startups is the pursuit of supply chains that reduce dependence on Chinese suppliers. Westmag, short for Western Magnetics Company, manufactures motors and actuators for drones and robots specifically as an American domestic alternative, with a seed round led by Andreessen Horowitz. Noble Machines, founded by engineers from Apple, SpaceX, and NASA, maintains two parallel supply chains, one in Asia and another in the U.S. and allied countries, to reduce geopolitical risk.
This concern is not just about business. Amy Yin of defy.vc was direct when commenting on the startup 1Robot, whose clients include Samsung and Nestle: our future wars will be fought with humanoid robotics. It is a strong statement, but one that sums up the climate of strategic urgency surrounding the sector. Companies like Machina Labs, which builds what it calls a software-defined factory capable of producing missile components in the morning and automotive parts at night, have already closed contracts with giants like Lockheed Martin.
Where innovation is heading
The picture that emerges from this list is broad, diverse, and quite revealing. Robotics has moved beyond being a technical niche to become one of the central bets of global venture capital. And this is not happening because investors suddenly got sentimental about machines. It is happening because the fundamentals have changed: costs dropped, technology matured, demand exploded, and the geopolitical context created structural incentives for this to happen especially in the United States.
The 25 companies mapped in this curation represent different bets on how this future will take shape. Some will dominate specific verticals, like GrayMatter, which automates surface finishing on products ranging from guitars to military fighter jets. Others will become horizontal infrastructure, those invisible platforms that power dozens of other products, like Foxglove and Dexterity. Companies like Zipline, which has been delivering medical supplies by drone since 2016, show that part of this future is already working today. And some, perhaps, will define entire categories that do not even have a name yet.
What is clear is that innovation in the physical world is accelerating at a pace that few sectors can keep up with. The combination of artificial intelligence with accessible hardware and real market demand has created a window of opportunity that the best funds in the world have already decided to walk through. Anyone following this movement closely knows that the next few years will be decisive for understanding which companies will come out ahead and which technologies will actually scale beyond labs and controlled pilots. The money has already picked a side. 🚀
