roughly the same number of deals from last year but almost 2X in spend - 20B+ vs 11B Q1 2025.

the inflation is driven by the mega deals (31 at $100M+), largest Q1 number since 2022, when we had 77. For the record, in Q1 2025 - 15 were 100M+ and in Q1 2024 - 27 were 100M+. And yeah, that’s equity transactions only.

this Q we’ve gotten record seed size closings in Europe - money above includes a $1B seed publicly announced (AMI) but doesn’t count another one that was never made public (Ineffable) as we don’t put rumours in the books.

AI and defence is where’s at, and the head of the market is as competitive and concentrated as ever - limited qualitative supply overly matched by top spenders doing a lot of creative gymnastics to sneak into deals. Bar is also very high, and what in previous years looked like a decent biz ($5-10m ARR, 2-3x yoy) now is being dismissed - different 2026 reality, w/ high opportunity cost for investors. American money is everywhere you look around, and smart managers un-willing to compete on deep pockets basis are already building deeper moats at the margins.

apart from those, the long tail tells a different story though. Beyond the glam of a few selected high-end performers such as Legora or ElevenLabs, we’ve gotten a rather grim picture, with many marginal, iterative and/or non-imaginative point of views. There’s a huge gap between what you can do with AI and what value propositions are out for sale - early stage dealflow shows that Europe badly needs vision-driven people with big balls of steels, but what else is new, that’s the story of a whole continent, right?

the question is, of course, not how many Linkedin scrappers sold as AI sales companions can be built - that’s for the market to work out, but how many of those deserve risk capital financing. On one hand, the output reflects the actions of its actors taking what they’re being given, but on the other alpha doesn’t grow on trees - real talk is that on a deeper level it’s rather hard to distinguish distinct deployment strategies beyond the tier one, which is a pretty thin circle of people. Same like in any industry, I guess - and of course, that’s just the half empty side of the glass.

a lot of fresh powder this quarter - we have tracked about 100 new funds that made public their market availability. Three buckets here:

Big spenders are the usual suspects:

Multi-stagers

Up and coming

Pan-EU

Local champions

Other stuff

All these quick extracts from our data coverage over at Nordic9 - give it a try already, it’s great value for money.

OpenAI has paused its Stargate UK data center project, the massive AI infrastructure scheme to deploy up to 8,000 GPUs in partnership with Nscale and Nvidia - the UK is not the natural home for many gigafactory-scale DCs aimed at huge training runs (heard a related conspiracy theory involving Starmer and Trump but we don’t run this kind of stuff around here)

Meanwhile in France Mistral took over 800 million in debt for a similar infra play, namely building a Paris-based DC to provide a 13,800 Nvidia chips sovereign cloud that competes directly with Azure and AWS for European government contracts.

The Brits have officially cut the VCT tax relief from 30% to 20% - the last time it happened (2006-2007), fundraising for VCTs dropped by two-thirds. Alas market was imature too back then.

76% of all European exit value in Q1 of 2026 came from secondary buyouts.

The sovereign play of the week comes from Spain, where the government paid McKinsey $500k to help them invest money in their local startups.

This week’s Isar Aerospace second qualification flight of the Spectrum rocket from Norway was postponed due to a leak in a composite overwrapped pressure vessel one hour prior to launch. The first test flight took place a year ago but ended in failure after 30 seconds also due to tech reasons - the mission aims to achieve the first orbital flight from continental European soil and carry six payloads, including five CubeSats and one technological experiment.

Tesla got its first (supervised) FSD approval in Europe - in the Netherlands.

An explanation in simple words why European hyperscalers stand no chance of being significant global players.

This sounds like an anthropology experiment but it’s for real - while Europeans bought (again) huge volumes of Russian LNG in March, they continue to push the de-Americanisation of their tech consumption.

Euro area annual inflation rose to 2.5% in March 2026, up from 1.9% in February, driven primarily by a sharp 4.9% increase in energy costs. In the US, consumer prices were up 3.3% by comparison. The fun should start later this month.

More macro - the US economy grew by just 0.5% in Q4 2025, big drop from 4.3% of Q3 2025 and a bit different than ‘the highest growth we have ever had’ claimed by the American president at the time.

Frame of the week - Donald Trump … has threatened to destroy a civilization. How does an investor process that? Is it a bigger upside risk or downside risk?

Creandum’s big bet on crypto

the heterogeneous computing path from Europe

European data centers in space

bear market in Europe

Europe’s most active angels

who is doing nuclear business in Europe atm