Deeptech in the Netherlands in 2026: quantum, chips and autonomous systems attract investment

From quantum processors in Delft to brain-inspired chips in Rijswijk: the Dutch deeptech sector is showing in 2026 a broad spectrum of companies building foundational technology. In total, around 42 startups are active in this segment, with a clear concentration in hardware, AI chips and autonomous systems.

The funding rounds closed over the past year vary considerably: from a seed round of €7 million for a lab software company to a Series A of €400 million for a drone and turbojet company. That spread reflects how diverse the sector is, but also how unevenly capital is distributed across the different technologies.

The largest rounds go to hardware and defence

The largest round in this overview belongs to Destinus from Katwijk: the company raised €400 million in a Series A in early 2025. Destinus develops AI-driven drones and turbojets for the defence market. The fact that such a substantial round goes to a company founded only in 2024 says something about the momentum that defence technology currently has among investors.

Planck Network from Harderwijk ranks second with €200 million in a Series B. The company is building a decentralised GPU network for training and deploying AI models. The underlying idea is that computing power can be distributed more broadly than through a handful of large data centres.

General Intuition, founded in 2025, closed a seed round of €114 million. The company trains AI agents on billions of gameplay images to improve spatial reasoning. CuspAI from Amsterdam raised €111 million in a Series A for a platform that searches for and generates new materials, claiming to do so ten times faster than through traditional methods.

LUGN Security from Heerlen, active with sensors, drones and AI for monitoring large sites and infrastructure, closed a Series B of nearly €37 million. Axelera AI from Eindhoven received more than €61 million via an EU subsidy for energy-efficient AI chips targeting computer vision at the network edge.

Quantum and edge computing as distinct clusters

QuantWare from Delft represents the quantum cluster. The company supplies scalable quantum processors to quantum computer builders worldwide and has chosen the role of supplier: QuantWare positions itself as a manufacturer of the underlying hardware, rather than as a provider of complete quantum systems. With €27 million in an extended Series A, the company has room to develop that strategy further.

At the other end of the spectrum are companies that aim to make computing power as small and energy-efficient as possible. Innatera from Rijswijk builds brain-inspired chips that operate intelligently at under one milliwatt of power. The computing power is thus located close to the sensor itself, which is relevant for applications where bandwidth or battery life is a bottleneck. Axelera AI is working on similar principles, but specifically for computer vision.

Spectro-AI from Enschede combines autonomous AI with drone inspections in the field. The company has been active since 2018 and closed a Series B of €19.3 million in 2022, making it one of the more mature players in the Dutch deeptech landscape.

Materials science and data infrastructure as emerging themes

CuspAI focuses on a field that rarely receives attention: designing new materials using AI. Materials development takes a long time and costs a great deal of money. AI can accelerate the process. Applications range from batteries to construction materials. CuspAI was founded only in 2024 and has already attracted €111 million, indicating that investors take the potential of this segment seriously.

Roseman Labs from Utrecht is working on a different type of infrastructure: collaborating on sensitive data without any party being able to view another's information. With €9.4 million in a Series A, it is a smaller round, but the theme, privacy-preserving data analysis, is becoming more relevant as organisations increasingly want to share data without incurring legal or ethical risks.

Balthazar from Amsterdam is building an operating system specifically for R&D labs in deeptech. It addresses the problem that labs lose a great deal of time on administration and work with fragmented tools. The seed round of €7 million is modest, but the company is targeting a concrete bottleneck within the sector itself.

Structural challenges in the sector

Deeptech companies generally have longer lead times than software startups. Developing, certifying and manufacturing hardware requires more time and capital than launching an app. This is reflected in the funding patterns: rounds are large, but companies such as Innatera and Spectro-AI have been operating for years before reaching a significant position.

The way defence and security applications are becoming intertwined with civilian technology is a theme that affects several companies in this overview. Destinus and LUGN Security explicitly target that market. This requires thorough certification and close coordination with governments and procurement processes, which further extends the commercial cycle.

The geographical spread is notable: the companies are located in Delft, Eindhoven, Rijswijk, Enschede, Heerlen, Utrecht, Harderwijk and Amsterdam. There is no single dominant hub. That can be a sign of strength, but it makes collaboration and the flow of talent between clusters more difficult.

For investors and policymakers in the Netherlands and Europe, the combination of large capital requirements, long development trajectories and the strategic importance of the technologies is a point of concern. The rounds being closed now are largely sourced from international investors. Whether the European financing infrastructure, including EU subsidies such as that received by Axelera AI, is sufficiently aligned with the scale at which these companies operate is a question that is being raised more frequently in policy circles.