Inside the Swiss Deep Tech Report 2026: Switzerland Out-Invests China and the US in Deep Tech

Which country puts the largest share of its venture capital into deep tech? Not the United States. Not China. \ The Swiss Deep Tech Report 2026 , published today by Deep Tech Nation Switzerland with Founderful , Kickfund , Startupticker.ch and Dealroom.co and launched at VivaTech in Paris, puts Switzerland first in the world. The rest of the data explains both why that lead is real and where it runs out. \ Deep tech is the slow part of the technology economy. It is the advanced computing, robotics, artificial intelligence and life sciences that need years of laboratory work before they earn a dollar. The companies built on it now sit at the top of global market-cap tables, which is why the question of where the next cohort is engineered carries weight. \ This is not one boom year. Annual capital into Swiss deep tech has climbed from $543m in 2015 to a record $2.6B in 2025, close to a fivefold rise, passing the previous peak of $2.5B set in 2022. The line is not smooth, 2023 dipped to $1.4B, but the decade trend is one direction, and it was built with effectively no public venture money. \ Switzerland leads, and not narrowly Between 2020 and 2026, 63% of all Swiss venture capital went to deep tech , the highest national share measured. That is ahead of China at 56% and the United States at 54%, against a world average of 43% and a European average of 34%. One caveat belongs in the open: this report folds life sciences into deep tech, a broad definition, which is part of why the figure sits above the roughly 40% that Dealroom's narrower European cut reports . Even on the conservative measure Switzerland leads. \ Intensity backs the share. At $1,470 invested per head, Switzerland commits more to deep tech per capita than any country in Europe, ahead of Sweden at $1,328, and ranks in the global top three alongside Israel and the United States. \ The pipeline is tilting to AI and robotics \ What gets built is changing. AI and machine learning have gone from 11% of newly created Swiss deep tech companies in 2010 to 2021 to 25% in 2022 to 2026, now one in four and the second-largest segment behind biotech. Robotics nearly doubled its share, from 5% to 9%. Biotech and techbio remain the largest single segment but eased from 31% to 26% as the pipeline diversified. \ Robotics is the sharpest relative lead. Switzerland created 3.8 venture-backed robotics startups per capita since 2020, against 1.1 in the United States and 0.8 in the UK. That is five times the UK rate, about seven times Germany, and 3.5 times the US. \ \ A decade ago Switzerland's deep tech identity was biotech, a sector with long timelines, patient capital and a domestic investor base that understood it. The 2022 to 2026 cohort is led by AI and robotics, which compounds two pressures at once. These are the sectors with the steepest capital requirements at scale, and they are the sectors where the deepest pools of growth capital sit in the United States rather than Switzerland. So the pipeline is not just diversifying, it is rotating toward exactly the kind of company that outgrows domestic financing fastest. The robotics lead sharpens the point: building 3.8 venture-backed robotics startups per capita is a genuine origination advantage, but robotics is hardware, and hardware burns cash through scale-up far harder than the biotech and software that defined the previous cohort. Switzerland is getting better at creating the companies that will most test the financing gap the report flags elsewhere. \ The engine is two universities ETH Zurich and EPFL rank first in Europe for new deep tech spinout creation, with 24 and 16 venture-backed spinouts since 2023, ahead of Oxford and Cambridge. Measured cumulatively since 2010, ETH and EPFL sit at 152 and 151 , essentially tied with Cambridge at the top of the European table. The report frames the Zurich-Geneva-Lausanne axis as the Alpine Tech Cluster, with more than 1,500 venture-backed deep tech startups, one of Europe's two superclusters alongside the London-Cambridge-Oxford corridor it labels New Palo Alto, at more than 3,400. \ \ The concentration is the strength and the fragility in equal measure. Two federal institutes, 30 minutes apart, out-create Oxford and Cambridge on new spinouts and effectively anchor one of only two deep tech superclusters in Europe. That density is why global funds now treat Zurich and Lausanne as a single destination worth flying to, and it is close to unrepeatable, because it rests on decades of research funding and a tech-transfer system that lets labs spin companies out cleanly rather than litigating ownership for years. The risk is the mirror image of the advantage. An ecosystem that runs through two institutions is exposed to anything that touches them, from Switzerland's research access to European programmes to the quiet competition from US and Gulf universities now paying to pull the same talent. The Alpine Tech Cluster is a real moat today. Whether it stays one depends on Switzerland protecting the two pipelines that feed it, because almost everything downstream, the founders, the spinouts, the international capital, traces back to them. \ Where the gap sits The most useful number in the report is the one that exposes a weakness. Foreign investors supply 88% of Swiss deep tech capital at rounds of $100m and above, with US funds alone accounting for 54%, against 75% foreign across Europe. Domestic investors provide 36% of the smallest rounds but only 12% of the largest. The companies are Swiss. The cap tables, once they scale, are not. \ Read the dependence across the round sizes and it is not a single gap but a gradient. Domestic capital holds 36% of seed and early rounds under $15m, slips to 18% in the $15m to $100m band, and falls to 12% once cheques pass $100m. Foreign capital runs the mirror image, at 64%, 82% and 88%. US funds move in the opposite direction to Swiss ones, climbing from 12% of the smallest rounds to 54% of the largest. The pattern means Swiss investors are most present exactly where risk is highest and ownership is cheapest, and least present where valuations harden, board seats are allocated and the bulk of the eventual return is priced. By the time a Swiss deep tech company is worth backing at scale, the people setting its terms are mostly sitting elsewhere. \ What the numbers actually argue My read is that the 63% headline is the softest figure here, because it moves with the definition. The findings that do not move are the spinout pipeline and the research base that feeds it, both expensive and slow to replicate. That is the moat, not the share number, and it is why global funds now arrive in Zurich on their own rather than waiting to be pitched. \ The trade, if there is one, sits at late stage. A market that funds 12% of its own $100m rounds is not a sign of weak companies. It is a sign of absent local capital meeting a cohort that is about to need a great deal of it, which the report itself flags as an opening for larger domestic tickets. Switzerland has spent a decade proving it can originate deep tech. The next two to three years will test whether anyone at home is willing to own the part that pays. \ Folding life sciences into deep tech is what lifts Switzerland to 63% and a record $2.6B; strip biotech and medtech out, as Dealroom's narrower European cut does, and the share settles near 40% with funding closer to $1.3bn. Both can be true at once, and the useful point is that Switzerland leads on either measure, which is rare. A number that survives a change of methodology is a number worth building a thesis on. A number that depends on a generous definition is a number to cite with care. The 63% is the headline; the durable claim underneath it is that deep tech is the default destination for Swiss capital regardless of how the category is drawn. \ The harder reading sits in the order of the findings. Origination, talent and company creation are leading indicators, and on those Switzerland ranks first or near-first on almost every metric the report measures. Capital depth and value capture are lagging indicators, and on those it sits fifth in Europe by absolute funding and among the lowest for domestic late-stage participation. The report is therefore a portrait of an ecosystem that has solved the slow, compounding part of the problem, the research, the spinouts, the founders, and has not yet solved the part that converts those into owned outcomes. That sequencing is normal for a young deep tech market, but it is also the precise reason the next phase is the one that counts: the inputs are proven and abundant, and the binding constraint has moved downstream to who finances, and ultimately owns, the scale-up. \ The full report is available here . \ Final Thoughts Strip away the rankings and the report is making one quiet claim: that a country can lead the world in deep tech without owning much of what it builds. Switzerland originates the companies, trains the founders, files the patents and produces the spinouts, and then watches the expensive half of the lifecycle, the growth rounds, the board control, the eventual exit, get financed and largely captured abroad. Deep Tech Nation Switzerland presents this as arrival. I read it as a choice that has not yet been named, and the more honest question is whether that choice is a position of strength or a polite form of dependency. \ The reason this matters now, and did not a decade ago, is that deep tech breaks the economics software taught everyone to expect. In software you could originate a company cheaply and scale it cheaply, so the place that invented a product usually kept it. Physical AI, robotics and advanced compute do not work that way. Origination is cheap and scaling is brutally capital-intensive, and the capital that funds scaling sits in a handful of mostly American pools. That structural split is exactly what the Swiss numbers describe: a country that is first in the world at the cheap, compounding part of the problem and fifth in Europe at the expensive part. Switzerland is not failing to scale its companies. It is running into the physics of who finances deep tech at the frontier, and it is the cleanest example anywhere of origination and ownership coming apart. \ For a few years this is a comfortable bargain. Foreign capital brings pricing discipline, sector expertise and global market access, and a small economy with no state venture arm gets to host a world-class ecosystem without underwriting its riskiest rounds. The catch is that the bargain has an expiry date, and the report quietly sets it. The 2023 to 2025 cohort is the largest Switzerland has ever produced, it is tilted toward the most capital-hungry sectors in the report, and it reaches the Series B and growth stage over the next 24 months. That is the window where the country either builds domestic muscle to own the upside or formalises its role as the world's deep tech research lab, brilliant at invention and content to rent out the returns. \ My view is that the origination lead is real and close to unrepeatable, because it rests on decades of research funding, a clean tech-transfer system and a neutrality that makes Switzerland a place global talent and capital are happy to meet. But a lead in inputs is not the same as a durable economic position, and the report's own structure, inputs first and abundant, ownership last and thin, tells you which way the value is currently flowing. The interesting story over the next two years is not whether Switzerland stays number one on share. It will. It is whether a single credible domestic growth fund emerges before the foreign-owned scale-up pattern hardens into permanence. Win that, and deep tech nation means a nation that owns deep tech. Lose it, and it means a nation that incubates it beautifully for everyone else. \ Don’t forget to like and share the story! :::tip Vested Interest Disclosure: HackerNoon has reviewed the report for quality, but the claims herein belong to the author. #DYOR. ::: \ \