The Deep Tech Investors Writing the Most Interesting Cheques in 2026

Photo by RDNE Stock project via Pexels

Deep tech now accounts for 36% of total global VC funding, up nearly three times since 2016, and the pace is still accelerating. U.S. private investment in computing equipment alone has roughly doubled since 2023, hitting $280 billion on an annualised basis in 2025. That is not a software-driven cycle dressed up in hardware clothing. It is a structural reallocation of capital toward firms building at the physics layer, the biology layer, the materials layer. The investors writing the most interesting cheques right now are not chasing the same AI application plays that defined 2023 and 2024. They are betting on the infrastructure underneath everything else, and their portfolios reflect a level of conviction that most generalist funds simply cannot match.

The context matters here. Deep tech venture has historically been penalised by its own timeline. Hardware takes longer to iterate, science takes longer to de-risk, and the commercialisation gap between breakthrough and revenue has scared off allocators trained on SaaS multiples. What changed after 2022 was not just ChatGPT. It was the sudden, painful visibility of hardware bottlenecks in an AI-first world. Data centres, energy infrastructure, advanced semiconductors, robotics, synthetic biology, all became obviously critical at the same moment. Generalist funds noticed. But the specialists, who had been building conviction and domain expertise for a decade, found themselves in the rare position of being both right and early at exactly the right time.

The firms showing up most consistently at the frontier round level are Lux Capital ($7 billion AUM, currently deploying from its $1.5 billion Fund IX), DCVC ($4 billion AUM), Eclipse Ventures ($4 billion AUM), and Khosla Ventures, which is currently raising a $3.5 billion vehicle. Flagship Pioneering sits at $14 billion AUM and has essentially built a model of internal company creation rather than external investment, producing Moderna as the canonical proof point of that thesis. Prime Movers Lab ($1.2 billion AUM) has carved a specific niche around what it calls breakthrough science commercialisation, targeting companies that can move the needle on energy, transportation, infrastructure, agriculture and security. Playground Global ($1.2 billion AUM) focuses on the full hardware stack, from silicon to systems, and has backed companies working at the intersection of AI and physical world robotics with notable discipline around capital efficiency. The Engine, MIT's affiliated fund with $1 billion AUM, continues to de-risk the earliest and most technically uncertain bets in ways that no purely commercial fund can replicate.

The investors who matter in deep tech are not the ones who showed up when the returns looked attractive. They are the ones who stayed when the timelines looked impossible.

What separates the most interesting investors from the merely active ones is not fund size. It is thesis clarity and operator depth. Khosla Ventures has long maintained that founders working on civilisation-scale problems deserve patient, technically literate capital, and its portfolio spanning energy, AI hardware, biotech and climate reflects that consistency. Eclipse Ventures has built a distinct identity around industrial transformation, specifically the modernisation of physical industries through software-enabled hardware, and its partners include former operators who have actually run factories and supply chains. Lux Capital takes a different posture, one that is more philosophically curious and deliberately contrarian, funding work in synthetic biology, neuroscience and advanced materials that sits outside any obvious market map. These are not interchangeable choices for a founder. The firm you take money from in deep tech shapes your roadmap, your hiring, your manufacturing partnerships and often your regulatory strategy. That operational alignment is worth more than a marginal valuation difference at Series A.

For founders, the practical implication is this: the deep tech funding landscape in 2026 is more stratified than it appears from the outside. At pre-seed and seed, the most valuable partners are often not the brand-name funds but the hands-on incubators, SOSV's IndieBio and HAX programs, university-linked vehicles and government-backed programs that can provide lab access, regulatory navigation and technical credibility before a startup has revenue or even a prototype. At Series A and beyond, the jump to institutional capital from DCVC, Lux or Prime Movers Lab requires a different kind of pitch, one built on de-risked science, a clear path to manufacturing scale and a founding team that has thought seriously about unit economics at volume. Founders who confuse deep tech VCs with generalist funds, and pitch accordingly, are leaving conviction on the table.

The next 12 months will see further concentration at the top end of deep tech venture. Khosla's $3.5 billion raise, if closed as expected, will give it significant firepower to double down on energy and AI infrastructure plays at a moment when both sectors face hardware-constrained bottlenecks. Defense tech, already at record funding levels, will attract more crossover capital as geopolitical risk stays elevated. The funds that built genuine domain expertise before the cycle turned, rather than pivoting to deep tech because the returns looked attractive, are the ones worth watching most closely. In a category defined by long development timelines and brutal technical risk, pattern recognition built over a decade is the only moat that actually holds.