Zwei Unternehmen, die andere in der Nacke halten, haben 10 Milliarden Yuan in ein Super-Unicorn investiert.
Some time ago, there was a huge wave of "learning the 996 work culture" in the European venture capital circle. The formation process of this wave and the debates between the two sides were extremely exciting and full of dramatic tension. I won't elaborate on them here. Friends who are interested can jump to the article "Foreign investors start to praise the 996 work culture". What you need to know here is that this was not just a carnival among ordinary netizens, but a serious discussion based on methodology among a large number of star entrepreneurs and top investors.
For example, Nik Storonsky, the founder of Revolut, the unicorn company with the highest valuation in Europe, Harry Stebbings, the host of the top - tier venture capital podcast 20VC, and Martin Mignot, a partner at Index Ventures, the venture capital firm with the best performance in Silicon Valley this year, all openly and unambiguously stated their positions in support of entrepreneurs working harder.
Among them, Harry Stebbings made the most severe and unsparing statement: "This is the biggest problem in Europe. If you keep saying you want to build a company worth $10 billion but work from 9 to 5, five days a week, you're just deceiving yourself."
The opposing side can basically only refute from a moral perspective. Because in the past few years when artificial intelligence has dominated the venture capital field, Europe has been a "lost land". Phenomenal products like Deepseek, ChatGPT, Claude, and Manus either emerged in the United States or in China. Even venture capital, which pursues non - consensus, has made a "decision against tradition" and bypassed Europe. According to statistics, the total scale of artificial intelligence venture capital in the European market in 2023 was $8 billion, far behind the $68 billion in the US market and also significantly behind the $15 billion in the Chinese market.
In 2024, the gap narrowed, but it didn't really matter. In that year, the scale of artificial intelligence venture capital in the European market was $11 billion, while the US market received $47 billion in the same period. Meanwhile, in an environment lacking investment themes, venture capital showed a high degree of concentration. The artificial intelligence track took away 33%, making the already under - resourced European venture capital circle even more difficult. The total number of venture capital events throughout the year barely exceeded 4,000, dropping to the lowest level since 2015.
In this vicious cycle, even Mario Draghi, the former President of the European Central Bank and the former Prime Minister of Italy, couldn't help but say dejectedly in a report: "Technology should have been the pillar of Europe, but now it has become the weakest link. This is by no means a coincidence."
However, recently, the European venture capital circle may finally have something to be proud of because of a milestone financing event: Recently, the AI unicorn Mistral AI completed its Series C financing with a scale of 1.7 billion euros (approximately 14.2 billion yuan), and its post - investment valuation reached 11.7 billion euros (approximately 97.8 billion yuan), becoming another large - model startup company valued at over $10 billion after OpenAI, Anthropic, and Perplexity. However, this is not the highlight of this round of financing. The highlight of this round of financing is that its investors include DST, A16z, General Catalyst, Index Ventures, Lightspeed, NVIDIA, and a name that Chinese people are both familiar with and have complex feelings about, ASML.
(Mistral's homepage)
Why does the lithography machine giant invest in AI?
According to current public information, ASML not only participated in this round of financing but also led it. Out of the total 1.7 - billion - euro financing, ASML alone contributed 1.3 billion euros (approximately 10.868 billion yuan) in exchange for an 11% equity stake. That is to say, this entire investment event can be described as an unbreakable alliance between the most representative European technology giant and the unicorn enterprise with the highest valuation and the most promising future in the European artificial intelligence track.
For the European venture capital circle, this is a dream combination.
European technology media have also generally given quite exciting evaluations. For example, a Yahoo commentator wrote: "This cooperation highlights Europe's efforts to break free from the technological influence of the United States... The increasingly hostile attitude of former US President Donald Trump towards EU technology regulation has sparked discussions about whether the European continent is overly dependent on US technology companies." Clara Chappaz, the French Deputy Minister for Digital Affairs and Artificial Intelligence, cheered on her personal social media: "European technological sovereignty will be established through you."
But as Song Dandan would say, these are just emotions. ASML's investment in Mistral AI is indeed worthy of commemoration, but as the birthplace of the market economy and an established capitalist region, Europe can't be unaware of the importance of "business is business". And "business is business" means that if ASML's investment is based only on the public opinion atmosphere and political correctness rather than its own business needs, this investment is likely to end up with an unsatisfactory result.
So after calming down, many people began to ask a question: Why does ASML invest in Mistral AI?
According to the press release, the reason why ASML and Mistral AI were able to finalize the cooperation is that they share the same vision, both hoping to tap greater value in the industrial manufacturing field. Mistral AI said that after receiving this 10.8 - billion - yuan investment, it will mainly invest in "customized decentralized cutting - edge artificial intelligence solutions" to solve the most complex engineering and industrial problems, aiming to "empower enterprises, public sectors, and all industries to enhance their competitive advantages" through advanced models, customized solutions, and high - performance computing infrastructure.
ASML said that Mistral's professional capabilities in AI will help ASML perfectly combine its leading capabilities in overall lithography products and services with their respective unique advantages. Christophe Fouquet, the President and CEO of ASML, further explained on his personal social media: "What Mistral offers is not just a large model. More importantly, it can help us develop better tools and solutions for our customers and gradually improve our own operations."
This seems to be a reasonable explanation, but it doesn't hold up to closer scrutiny.
First of all, although Mistral AI is one of the large - model unicorns with the highest valuation at present, its products are not popular in the market. A recent survey report released by the University of Amsterdam showed that Mistral's market share in the entire large - model track was only 2%, and it quickly fell behind Deepseek and OpenAI after January 2025. This means that Mistral has not shown convincing performance in terms of product differentiation or user experience.
(Source: The official Substack of the University of Amsterdam)
Even worse, Mistral AI currently doesn't have a good way to bridge the widening gap. As mentioned above, Mistral AI's current vision is to deeply integrate into Europe's developed industrial system and become an integral part of the manufacturing industry. Arthur Mensch, the founder of Mistral AI, also clearly mentioned this point in an interview with Time magazine in May 2024, saying: "Our main direction is to ensure that our technology can improve productivity and bring reasoning capabilities to certain vertical industries and fields, thus bringing social benefits."
However, the problem is that regardless of the loss of European local industries, industrial automation has been developing for many years, which means that the production processes in the manufacturing scenario are quite mature. The sunk cost for users to upgrade is high, and accordingly, Mistral AI's difficulty in acquiring customers has also increased. Under this premise, Mistral AI's revenue sources are currently very single, mainly coming from large contracts worth hundreds of millions of yuan and lasting for three to five years from "a small number of large customers".
Whether these contracts can be successfully fulfilled and their future sustainability are both unknowns. The only thing that is certain is that Mistral AI doesn't have much energy to recruit enough talent and resources in the context of the AI arms race.
Therefore, more and more people now believe that Mistral AI's current high valuation may be a product of "political correctness". An important reason is that Cedric O, one of the co - founders of Mistral AI, once served in the Macron government as the French Secretary of State for Digital Affairs. The influence brought by this position cannot be underestimated.
Another example is that Mistral AI actually started this round of financing in July. Rumored investors included the wealthy Abu Dhabi MGX National Fund, and there were also rumors that Mistral AI might be acquired by Apple. However, in these investment rumors, Mistral AI's valuation generally remained at around 10 billion euros, and it was said that the funds would mainly be invested in the commercial promotion of the chatbot Le Chat and the continuous development of its large - language model, which is quite different from the publicity after ASML's investment.
In addition, although ASML has made many external investments in recent years, with an annual investment budget of 2 billion euros for improving R & D capabilities, it has not made many equity investments (not to mention investing in trendy industries with a lot of bubbles). In recent years, the companies it has invested in, such as Cymer and SMART Photonics, are all in the semiconductor industry chain, consistent with its core semiconductor equipment business, especially focusing on the production of advanced lithography and photonics technologies.
In short, as the well - known current affairs commentary website Politico stated at the beginning of its report on this financing: The existence of these factors makes it hard not to suspect that this is a commercial transaction full of political overtones.
Work together to overtake on a curve
Anyway, having money is always a good thing.
ASML's investment in Mistral AI at least proves that "revitalizing Europe" is not just an empty slogan. The 1.7 - billion - euro investment can also, to some extent, help Mistral AI gain some strategic space and retain the possibility of transforming to survive. Some media reported that the completion of this deal marks that Mistral will shift to the development of industrial applications and abandon its consumer - facing chatbot business. And from the overall data, the European venture capital circles seem to have generally reached a consensus on this matter:
As mentioned before, in the era of venture capital dominated by artificial intelligence, Europe has fallen far behind the United States and China and is increasingly unable to compete on an equal footing in terms of resources such as talent and computing power. Under pressure, European venture capitalists are thinking outside the box and changing their strategies. They are now keen on investing in vertical - field artificial intelligence application products. According to statistics, as of the end of August 2025, the total venture capital in the European artificial intelligence application field reached $5.8 billion, accounting for 63.5% of all AI venture capital.
According to the research of statistical institutions, the general thinking of investors is that large - model development is too expensive and competitive. In contrast, "developing artificial intelligence applications for specific industries has a low threshold, usually requiring a small amount of data sets and manpower, which can reduce the overall cost and complete phased verification more quickly."
Breaking down the vertical application fields of artificial intelligence further, the healthcare sector is the most popular track among investors. As of August 2025, the investment in artificial intelligence applications in the healthcare field in the European market has reached approximately $2.1 billion, exceeding last year's total transaction volume. The projects cover areas such as accelerating drug R & D and analyzing medical data. The largest investment this year occurred in March, when Thrive Capital led a $600 - million investment in the British drug - R & D startup Isomorphic Labs. In June this year, Neko Health, a Swedish "AI + medical scanning" R & D company, also received a $260 - million investment from Lightspeed Venture.
In this atmosphere, ASML's investment in Mistral can be better interpreted as a part of "Europe's overtaking on a curve". There is definitely a factor of not being willing to admit defeat and boosting the morale of the European people, but it may also be the best solution for Mistral at this stage. Taking a step back to move forward and making the most of Europe's old industrial parks' last advantage is not a bad thing.
It is worth mentioning that Mario Draghi, the former Prime Minister of Italy who was worried as mentioned before, also put forward a proposal, hoping that the EU could double the budget of the "Horizon Europe" (the EU's core funding framework for supporting scientific research and innovation) to 200 billion euros and focus on more disruptive innovations such as artificial intelligence or semiconductors. He proposed to create artificial intelligence factories to train such models and increase resource investment in this field to reduce dependence on external markets.
It's really rare for Europe, which is often described as a "loose union" for a hundred years, to call for "everyone rowing the boat together" from top to bottom.
Of course, as mentioned at the beginning, Europe's biggest challenges are confidence and time. Because from the data, although Europe was not lacking in creativity between 2008 and 2021 and gave birth to a total of 147 unicorn enterprises, 40 of them moved their headquarters overseas, mostly to the United States. Just as the United States established the huge Bretton Woods system and the Marshall Plan back then and built the current market order in the 30 years after World War II, historical inertia is obviously not something that can be easily reversed by a 200 - billion - euro funding plan.
This article is from the WeChat official account "China Venture Capital", author: Pu Fan. Republished by 36Kr with permission.