Mistral AI acquires Emmi AI

334

ddoener 2 days ago 97 commentsRead Article on emmi.ai

HI version is available. Content is displayed in original English for accuracy.

⚡ Community Insights

Discussion Sentiment

83% Positive

Analyzed from 2577 words in the discussion.

Discussion (97 Comments)Read Original on HackerNews

MeteorMarc•2 days ago

Nice, also note that ASML is a big investor in Mistral AI, which made the industrial AI ambitions already more credible. https://www.asml.com/en/news/press-releases/2025/asml-mistra...

kergonath•1 day ago

ASML is the flashy, public-facing partner. Mistral is also working with the French government and defense industry for applications that are unlikely to be publicly announced, but are bound to bring in much more money.

sravanipuuta•1 day ago

The European-sovereignty angle is what makes Mistral's strategy harder to read from the outside. If a meaningful share of revenue comes from EU government and defense contracts, the public benchmarks against OpenAI/Anthropic stop being the right scoreboard — they're optimizing for different procurement rules and a different definition of "trustworthy." Curious if anyone has visibility into how those contracts are structured: per-seat, on-prem, or something closer to hosted with sovereign keys?

kergonath•1 day ago

I don’t know much about the really secret stuff, but Mistral is known to help customers build their own infrastructure to deploy their models and handle confidential data securely. They are also building data centers for their own cloud. It’s difficult to have a clear picture, in most cases we know about the partnership without having all the details (e.g. with ASML, CMA-CGM, or HSBC, or even worse with government or EU institutions).

snowpid•1 day ago

I'd like to add that we have publicly the procurement from the US government for the US models. And I also guess the Chinese government uses their model providers as well. So I wouldnt see so strange anymore.

dainank•1 day ago

Mistral AI and the Luxembourgish government also have a strategic partnership: https://gouvernement.lu/en/gouvernement/stephanie-obertin/ac...

barrenko•1 day ago

I believe Mistral has a deal with the EU directly also.

pm90•2 days ago

can you explain how it makes it more credible? is the assertion that asml is using mistral as part of its research/manufacturing?

SyneRyder•2 days ago

From the link:

"...a long-term collaboration agreement to explore the use of AI models across ASML’s product portfolio as well as research, development and operations..."

ASML is one of the clients Mistral keeps referencing, for example here: https://mistral.ai/news/forge But it isn't clear exactly what they've been doing together. The Forge page only mentions they "train models on the proprietary data that powers their most complex systems and future-defining technologies."

spiderfarmer•2 days ago

ASML knows like no other the importance of doing research in secrecy.

mtct88•2 days ago

ASML is one of the bigger investor of Mistral

https://mistral.ai/news/mistral-ai-raises-1-7-b-to-accelerat...

john_strinlai•2 days ago

the assertion is that the people at asml are likely in a good position to assess ai use in complex industry

yieldcrv•1 day ago

It doesn’t

ASML’s EUV money printer has nothing to do with their ability to deploy that money in illiquid investments instead of to their own shareholders

ASML buying equity in one company in tangentially related industry (just because they’re in Europe and the pickings are slim in both offerings and growth capital) has nothing to do with any synergy or integration with ASML’s utility (and bottleneck) to the chip supply chain

remember when you were studying for standardized tests as a teenager? this is what the high scoring answer would be

stingraycharles•1 day ago

A friend of mine works for ASML and it’s suggested they were nudged quite heavily “from above” to make this investment, rather than it being an actual strategic play by ASML. Basically “sovereign EU AI” is the play here.

xnx•2 days ago

Abbreviated title leaves out key detail: "Mistral AI Acquires Emmi AI to Create the Leading AI Stack for Industrial Engineering"

joegibbs•1 day ago

That’s more realistic, I was going to say…

reenorap•1 day ago

Is Mistral still competitive? I completely forgot they existed because of how much press the Big 3 get (Google, Anthropic and OpenAI).

tugdual•1 day ago

They’re not competing in the same domain - if you look at their business model it actually is much closer to ML consulting for companies (CMA CGM, ASML, Airbus…). The big three are trying to capture B2C mainly while Mistral is full focused B2B

throwaway260124•1 day ago

What’s in Europe ( or maybe not to generalize in France/Germany) that creates this bias for consulting?

ben_w•1 day ago

I don't know about France, but here in Germany I think there's more room for B2B than B2C because of the desire for stability, itself leading to bureaucracy in everyday life that sets expectations for a much slower everything (like, buying a house took 7 months, many contracts have 3 month notice periods).

But that's just my best guess, and I'm saying this as one who migrated here rather than growing up here. I've also actually noticed the literal anarchists here, whereas the ones in the UK I only knew once they told me, before anyone makes a planet-of-hats kind of mistake on this.

port11•about 19 hours ago

An hypothesis from my own consulting: firing an employee can be somewhat difficult, and their salary has quite some taxes on top. At my previous startup we’d rule-of-thumb calculate someone as costing us 140% of their pre-tax salary.

Put these two together and a freelancer becomes an interesting proposition for some tasks. In Germany I could bill roughly 2.5x my after-tax salary. You also incur no other costs such as equipment and illness-as—a-loss. And… you’re really easy to fire when COVID comes around.

snowpid•1 day ago

It brings more money. Europe in general has much more industry (hence China shock might be bigger) so Industrial AI makes more sense. And also the amount of Consumers needing industrial consulting is not high.

antipaul•1 day ago

"The big three are trying to capture B2C mainly"

I think it's more complicated. Anthropic has been focused on enterprise for a long time, and OpenAI seems to be doubling down as well.

sbinnee•1 day ago

I check almost every mistral model. But to be honest after agentic coding has become a thing, I only test their models in a chat app. I tried their coding TUI a few months back and found that it was way behind. It didn't even support skills back then. I wish they go strong just because they are a big AI player in Europe and I have personal connection to France.

phillc73•1 day ago

Mistral’s Vibe CLI does support SKILL.md files.

lanthissa•1 day ago

mistral was never competitive and is getting less so, but that doesn't matter they cant be allowed to fail and have a long time to find their lane. They're smart and have an audience of like 600m people and the largest governments by spending who would use them if they were good enough.

Barbing•1 day ago

I wonder if they could’ve caught up to Qwen & Gemma by now by now distilling them?

If their best cloud-run offering is far more intelligent than the laptop Gemma / Qwen than nevermind

jazzyjackson•1 day ago

I would speculate that the purpose is not just to have a domestically owned model that’s as good as the Chinese one, but for the investment to be used to build a domestic industry of people that know how to do it themselves, so as not to rely on foreign parties, so it doesn’t have to be better, just has to be good enough.

sureMan6•1 day ago

If you're a European AI lab and you see you can't compete with the Amerikans on compute or spending, and the Chinese are open sourcing their LLM work and keeping the Amerikans honest, then there's no real need for you to focus on LLMs

maelito•1 day ago

I'm using Vibe on a regular basis, it's great. Not missing Claude code much. It's also getting better week after week.

matthiasrsl•1 day ago

I'm statisfied by vibe coding-wise, but i found their TUI UX to be abysmal (not that CC is great, but definitely better). I'm talking to the extent of some characters from the numpad not being typable in the prompt. Have you had a similar experience?

phillc73•1 day ago

I don't use the numpad characters, but I have tried Vibe, Goose (GUI and CLI), Dirac and the built in agent in Zed for vibe coding. I keep coming back to Mistral's Vibe. I actually find the ergonomics of it nicer than the others I've used so far. I really wanted to like Goose, and their GUI offering is OK for chat, but I thought their CLI was poor. Dirac was OK and I should try it again to be fair. Zed was just overkill and complex for what I needed. Vibe CLI seems to hit the sweet spot, although it's not perfect. The challenges I encounter are mostly down to API errors though and sometime bash tooling. I could configure it better for that, if I took the time (which I should).

Barbing•1 day ago

> Big 3

This is why Elon’s appealing. I thought they surely had the talent to be considered a fourth at this point. (Oh someone mentioned they’re politically unpopular at work.)

Mistral is super welcome competition, good luck!

jatora•1 day ago

No they really don't because Grok is not a competitor. The big 3 are the big 3 because they have historically traded the top place for model intelligence. Grok has never crested that high. People like to think Grok isnt as popular because of elon and politics, but if Grok was the best coding model, nobody would give a fuck. Google has also not led since gemini 2.5 pro. It's really the big 2 at this point.

reenorap•1 day ago

How much on Kalshi to bet that xAI screws over the Cursor folks out of their $10B/60B deal through some loophole?

Only desperate people would go work for Elon at this point.

impulser_•1 day ago

Not really. They have usable models, but it probably Anthropic, OpenAI, Google -> Chinese labs -> NVIDIA, AllenAI, Mistral.

holoduke•1 day ago

The are in Europe and in France. The worst possible situation to be successful as a startup. I mean the extreme hiarchy, lack of salaries and taxation is not really gonna help.

menaerus•1 day ago

And they obviously don't hire the right people. Reasons can be many-fold. One possible explanation is that there's not many talent left on the market, and most have been already picked up by other AI labs paying more $$$ while offering more exciting work and more exciting trajectory at the same time. Another possible explanation is that there is enough talent on the market left but their recruitment process doesn't allow them to recognize those people, hence it is broken. When I look into their job postings, I tend to give higher chances for the latter.

techsystems•1 day ago

Competitive in what sense? In training LLMs? Then they would be below Korea then UAE models.

kriro•1 day ago

If I'm only thinking about non-programming business applications, anecdotally, Mistral is certainly a player in the European enterprise market. For most German companies I have interacted with, Mistral was the first point of contact regarding corporate AI rollout. For the "small potatoes" day to day minutia Copilot is probably the #1.

ChemSpider•2 days ago

Built what you want to use yourself. AI for engineering and physics sounds like the perfect product a company like ASML (Mistral investor) could use.

SilverElfin•2 days ago

I was skeptical when I saw the headline. And I still am. But AI for manufacturing and industry seems like a good way to differentiate and focus on a vertical that others are ignoring.

What I am curious about is what has Emmi actually built? Who uses it? I was hoping to see something like a demo on the website but couldn’t find anything concrete.

mediaman•1 day ago

They built a transformer-based mold flow simulator. Mold flow analysis is used in injection molding to predict whether and how the plastic will fill a cavity. It's helpful to make sure the mold will actually be able to fill, and to try to engineer lighter weight or different geometry.

Mold flow simulators already exist, but they're slow. So if you're a mold engineer, and you want to try different material properties, each run takes time.

By making this into a transformer (no idea what that model looks like) it can run pretty fast, because it's not mathematically solving it, it's running a learned function approximator. So for the mold engineer, suddenly they can just change values around and get pretty fast analysis.

They appear to not be selling it as a product per se, but partnered with a German company called Simcon to sell it, whose website lists it as a preview, so it's unclear how good it is compared to conventional simulations.

KeplerBoy•2 days ago

They seem to be doing let's say case studies on how AI based simulations can help industry.

They did injection molding for example, and I'm sure they're testing similar approach to everything which can be modeled by PDEs, which is well pretty much everything ever of engineering interest (I'm assuming this is somehow connected to a research project funded by Engel, one of the leading injection molding machine companies, located in the same Region): https://www.emmi.ai/models/neuralmould

ChemSpider•2 days ago

"Early 2025: First enterprise contracts secured" "Today: Powering Fortune 500 engineering teams" - I guess that is all that is publicly available for now.

I_am_tiberius•1 day ago

I assume it's more a talent acquisition as the company only exists for 18 months or so and you can't acquire enterprise customers that quickly, not even as an AI vendor.

bradley13•2 days ago

I'm glad Mistral is doing well, but...

I am so tired of M&A. Buy instead of competing or - heaven forfend - cooperating.

smt88•1 day ago

The reason we have so much M&A is that large companies can’t really innovate, especially when they’re publicly traded and have shareholders who hate risk-taking

tormeh•1 day ago

Mistral is not large, though.

smt88•1 day ago

800+ staff across 30 countries, and headquartered in a notoriously bureaucratic country

cosmobiosis•1 day ago

Thta's amazing, Mistral is winning silently lol

jjice•1 day ago

How so?

shevy-java•1 day ago

Skynet slop is growing.

atleastoptimal•1 day ago

Europe won't be a leader in AI unless they get huge capital investment and datacenter build up. Sadly the regulatory environment of Europe combined with brain drain to America means it won't happen

Propelloni•1 day ago

Since 2024 the USA is draining brains into Europe in science in general but esp. in climate research, health research, and "tech" research. I can't find good numbers on IT specifically (except that the UK seems to be draining towards EU).

Commercialisation of research is still hard in comparison to the USA, though.

atleastoptimal•about 20 hours ago

I know that but it's about AI. AI progress = research x compute, without compute then you don't get progress or leverage.

joelthelion•1 day ago

I see it the other way around. In a few years, algorithmic and hardware improvements will likely make these huge datacenters mostly obsolete.

Americans will get to enjoy their rusty infrastructure and polluted air, and Europe will have lean and clean infra to support just what's needed.

atleastoptimal•about 20 hours ago

That's like saying that improved gas mileage and energy efficiency will make oil obsolete.

In certain domains there are no ceilings to return on compute. For example, offensive/defensive cybersecurity, it will be an arms race of who has more compute to patch up against more sophisticated attacks, and who runs more sophisticated attacks

>Americans will get to enjoy their rusty infrastructure and polluted air,

America has very high air quality and fairly good infrastructure given their population density.

joelthelion•about 9 hours ago

ICE Cars have been around for 100+ years. We have a good understanding of their achievable efficiency, including theoretical guarantees.

LLMs on the other hand have only been around for a few years. Large technological breakthroughs are much more likely.

In addition, I don't think the future is billions of people chatting with ChatGPT all day. LLMs can write deterministic code for many things, and in the end, we only need their "intelligence" and brittleness in relatively few scenarios. So, with good optimization, we shouldn't need so many huge data centers.

On the topic of clean air, the US is relatively spared at the moment because past governments were more reasonable. But just wait a couple of years under the current leadership and you'll see. Just from this morning, look at the top comments: https://news.ycombinator.com/item?id=48214017

Aldipower•1 day ago

Meanwhile American AI is trained continuously on European data folks willingly gift to the big three AI companies.

u8080•1 day ago

Yeah, right, like they've already done that with social networks and chipmaking.

drstewart•1 day ago

Exactly. The same way the US is winning the EV / solar race - let China build obsolete factories that require way more investments that will be obsolete soon because of automation, then swoop in to build just what's needed and dominate. China will be left with scarred ugly landscapes while the US wins, as you'll no doubt agree

joelthelion•1 day ago

AI isn't a necessity in the same way that EV / solar is, we can afford to take it slow and see what sticks. Also, EV/Solar are a much more mature technology than AI, so huge technological leaps are less likely.

impossiblefork•1 day ago

Yes, and this isn't possible in a sustainable without a hardware-software carousel, which means that we must get EU AI hardware firms, including training accelerators.

The present situation must be seen as a sort of "keep alive" state, where we do something unsustainable until we can achieve takeoff, but we can only actually start up for real once we start making the machines we need.

If NVIDIA's margins were 20% or something, this wouldn't be the case, but they aren't.

mejutoco•1 day ago

Europe is slowly moving to put part of pensions in the stock market (like USA does). Right now people put it in sp500 so that sweet pension money in eu markets might be interesting to see.

NookDavoos•1 day ago

I can't imagine being an American approaching retirement with 45-50% of my pension portfolio in AI-related stocks (that's the current weight of AI companies on the SP500). Besides, when SpaceX/Openai/Anthropic go public, they'll get another hefty chunk.

u8080•1 day ago

Isn't sp500 mainly US companies with <10% EU ones?

aborsy•2 days ago

It’s interesting that a French company can compete at international level to some extent, given the regulations, labor laws and generally the business unfriendly environment. I suspect they capitalize on the preference of European governments to use EU products, but might be wrong.

maelito•2 days ago

This comment is ridiculous. France's economy is bigger than 90 % of "unregulated" countries.

European regulations help protect from USA's tech monopolies. French labor laws and social security and state-funded scientific schools helped build one of the most competent international AI scientist generation.

All of europe got crushed by the US on the domain of internet. "Regulated" or not.

linkregister•2 days ago

Indeed, French labor laws and their downstream effects have pushed the most talented French researchers to US-based frontier labs, thus building one of the most competent cohorts of international AI scientists.

kergonath•1 day ago

This has more to do with the humongous amounts of money sloshing around in VC funds and the disproportionate importance of the US in the global financial markets. They just followed the money. Those who are successful in securing funding then tend to come back eventually.

olivermuty•2 days ago

Do you really believe this? Lol

Could you please elaborate what labour law drives the labour out of france?

holoduke•1 day ago

You would be surprised that many regulations are lobbied by American companies to secure their dominance. From gambling to advertising. European regulations are not to help citizens. But to help big institutions.

guerby•1 day ago

That's exactly what the CEO of Mistral Arthur Mensch says in this audition before french legislators:

https://www.youtube.com/watch?v=I20fcnkzj5A

retinaros•1 day ago

Its not ridiculous but accurate. You right the scientist generation is huge from europe but many leave europe… Brain drain in france is huge nowadays because of what op states.

And if you look at mistral biggest customers it would be lying not to say they are done through political ties. No shame in saying that. US gov and agencies created FAANG through those same mecanism

keybored•1 day ago

It’s predictable that a board full of 100K USD+[1] salary programmers and capitalists are negative against labor laws.

[1] Outdated by now (higher)?

pezgrande•1 day ago

I have yet to see someone recommending Mistral for anything tbh.

phillc73•1 day ago

Their Voxtral[1] speech models are really good.

[1] https://mistral.ai/news/voxtral

nmfisher•1 day ago

+1 for this, I've found Voxtral to be the best combination of price/speed/accuracy.

qrios•1 day ago

We are using it for "old-fashioned" use cases (sentiment, classification) for some clients here in Europe. Mistral Small 3.2 8bit is good enough for most well-defined cases.

It may just be greenwashing to check “AI sovereignty” off the list.