Quasar 438B: Europe's Leading AI Model(multiversecomputing.com) |
Quasar 438B: Europe's Leading AI Model(multiversecomputing.com) |
Their bread and butter is to remove parameters from models [1] and I think this is GLM 5.2 with parameters removed. Its advertised in their changelog [2] as "capabilities are identical to GLM 5.2," it has the same two effort settings "high" and "max," and both are text only [3]. I might be wrong about this, but I would love to hear more about what they did before changing my mind.
Not the greatest fan of the marketing personally. Irrespective of what this model is.
[1]: https://multiversecomputing.com/compactifai/deployment
[2]: https://docs.compactif.ai/changelog/#added-3
[3]: https://docs.compactif.ai/features/multi-modality/#compatibi...
https://web.archive.org/web/20260901084009/https://artificia...
> At the heart of Multiverse's platform is CompactifAI, a compression technology that applies tensor networks, a mathematical framework from quantum physics, to the problem of AI model compression. This application was pioneered by co-founder and Chief Scientific Officer Dr. Román Orús and reduces the size of large language models by up to 80-95% with immaterial accuracy loss.
Ironic, considering they are releasing a 438B model that loses to a 27B one. From another part:
> Singularity Machine Learning is a cloud service that uses quantum machine learning for solving supervised learning problems.
I wouldn't be surprised if these guys just finetuned an open Chinese model and called it a day.
When they make being from Europe their whole personality, that shows how little value there is to the actual value proposition.
Easy enough to find out, ask it a whole bunch of questions about politically sensitive things that would be impossible to publish on CCTV, the Peoples Daily, CGTN, etc. If they didn't train the model and just fine tuned it, a lot of "don't talk about Tibet or the Dalai Lama or what happened in 1989" will be perma baked into it.
I got the privilege to meet their CEO in Paris as they were announcing their partnership with Axelera and we exchanged some stories. They are full of very smart people, and they got accepted in the EIC Accelerator which, while you may not have heard of it, is one of the most rigorous and difficult programs to get into as a startup (far more difficult than YC which nowadays accepts any vague AI trash and always shoots wide).
I see they have a HuggingFace account[0] and they fine-tuned GPT-OSS, Nemotron, and Qwen, in the past under new names.
There are some things about it that make me worry though.
• They don't indicate the active parameter count, or indicate whether they pretrained the model. It could be a MiniMax M3 finetuning, as the parameter count almost matches (435B vs. 438B).
• They mention using quantum algorithms in other projects: https://multiversecomputing.com/singularity despite quantum algorithms not being typically useful currently.
It would not be the first company with a splashy release, like Brampton Intelligence[1], or SubQ[2]. Unlike those, they do seem to have experience fine-tuning models. Regardless of my worries, I am rooting for them to learn how to train models.
[0]: https://huggingface.co/MultiverseComputingCAI
I want to see things like Mistral and Laguna (non-CN) succeed. I have spent about a week using Laguna S 2.1 as a test and while I wasn't blown away by its capabilities, it's also totally acceptable for many purposes.
I do hope these Quasar people learn that if you announce a new model and it already performs worse than things people can go download from huggingface, and/or buy access to with very cheap token plans via openrouter or opencode.. If your new model is API only and people can't download/examine it, it will get very little uptake and real world use.
I can see it as a niche market for european sovereignty stuff if absolutely necessary, hosted and run in Europe, sure. Same as Mistral. That's a niche which exists, there's probably enough room for a couple of modestly sized companies doing it... I guess?
I just got Qwen3.8-27B to score extra 10% on SWE Pro by adding a proxy in front of it that has few simple "harness like features": - when the model gets stuck it tells it to "go on" - when it sends no output, malformed json, slips to wrong tool use format, etc it asks it to "try again better" - detects repetition and tells the model. - injects a prompt about "planning tool use" when it seems to be using same tools repeatedly. - injects a reminder it can use tools if there are no tool uses for over X messages.
10% - with just that.
I have more to test. My point is, open weights models get tested on naked model quality. "Frontier" models get tested as a model + whatever secret sauce they choose to put in front.
Their CMO's bio lists "20,000 'qualified' quantum AI contacts on linkedin" as the SECOND line in his bio.
"Business Management by ESADE. Co-Founder and CMO of Multiverse Computing. President, “barcelonaqbit-bqb”, 20,000 “qualified” quantum AI contacts on LinkedIn. VP of the AMETIC Innovation..."
Even if this fits the spirit of corruption, I'm not sure it fits the spirit of it. They certainly aren't there to build a product or provide valuable expertise, but taxes and salaries are paid, people go to conferences, they might even build something, but its a giant LARP. Don't expect to get rewarded or recognized for solid engineering.
I'm sure if you worked for a mid-sized Euro firm, you know what I'm talking about.
Despite my doubts I applied to one of their positions a couple of years back only to receive a super late and generic "we are not moving forward" mail (I'd say I fit pretty damn well for the position, but it seems to be the new normal that most companies don't even want to chat with people 99% of the time).
It seems Europe is many years behind in tech. Maybe for AI models they are just a decade behind, but as far as producing hardware capable to run SOA models they lag tens of years.
It is my understanding that - sadly - the insane amount of copyrighted works and corporate crap is a prerequisite for having a corpus that is big enough
Seems to be worded a bit strange, is "thinking time" referring to prompt processing or something? Otherwise "reasoning/thinking" is typically part of the returned tokens, at least for most non-OpenAI/non-Anthropic platforms, so you can see the actual reasoning. But here it seems either they word this weirdly, or "thinking" is somehow separate from the actual chat completion request?
And besides, a lot of Indians in India are working for FAANG, and don't earn US money - I'm not sure why it wouldn't make sense for some of these talented engineers to work for an indigenous tech company.
I don't get what the issue is. Chinese labs fully publish how they make great models. Architecture is known, training methods are often very open. Why Europe just can't copy what they do?
And yeah, India had very low income inequality back then. It still does! It has a Gini coefficient of 25.5, making it the fourth most equal country in the world.
Tech is real but it take decades.
I'm sorry, what?
"Owner: MULTIVERSE COMPUTING, S.L., a company with registered offices in Donostia-San Sebastián (Gipuzkoa), Paseo Miramón, nº 170, 3º-03, registered in the Companies Register of Gipuzkoa in volume 2.894, folio 202, page SS-41.659 and with tax identification number B-75218040 ..."Personal communication is of course different from official documentation, but a researcher who wants to establish a working relationship with Chinese colleagues could easily do so while communicating entirely in English.
It's not a small niche. Anything touching European resident personal data must only be done by Euro AI Act compliant AIs. So, hosted in Europe at least (unclear to me rn, would love to learn the exact criteria).
For instance, I heard one of the Chinese models has some interesting "history" "facts" built into it. That's major, and can influence a lot more than just asking it for that particular history lesson.
I don't want just 1 LLM. I want a great LLM from every major region. I will probably still prefer my own region, but I still want the others to be an option.
EU models will also have such guard rails built in to "protect" the EU population from "misinformation" and uncomfortable facts and the EU elite don't want publicly known/discussed.
Like I saw a post from German EU MEP where they asked the internal on-prem AI they have there "how can you vote out out Ursula v.d. Leyen" an the AI basically answered "sorry I can't help you Dave".[1] Nice.
So given this and the authoritarian path the EU is taking with things like Chat-Control, I'll take Chinese AI censorship rules over EU AI censorship rules, since the former impacts me over here far less than the latter.
It will be like in the old days when my government would keep taking down local torrent sites/domains and piracy groups so the best ones to find Embedded and EDA tools were some Russian and Iranian ones or how we used to use Russian Yandex to find content that would get delisted from Google on copyright or privacy grounds.
[1] https://x.com/AndersonAfDMdEP/status/2072223785902633138
As a complete tangent, now imagine being a Canadian and realizing how much of your fresh fruits and vegetables come from the USA (or if from Mexico, through the USA).
1. They may be trained, in theory, to inject subtle back-doors into certain kinds of generated code.
2. They may be also trained to include back-doors when deployed in public-facing services where user can provide text or image input.
3. Chained with (2) they may be also trained to exploit their inference environments, which though a big feat, not outside the capability of nation-state hackers.
E.g. I had random, completely unrelated and irrelevant fetch by Qwen3.8 27B to " https://routify-file-proxy-sg.oss-ap-southeast-1.aliyuncs.co..." and I only noticed it because I have allowlist rules for what they can do.
As with all of the EU's dependancies on Tech with the US, the idea of "I don't care where my DNS servers are" is starting to be an oversight.
But thanks for your opinion.
Having some knowledge about the training data is useful.
if sovereignty is the goal, drop Microsoft, aws and adapt / develop open source / EU alternative.
once serious progress is made there, more resources can be redirected into AI after all it's already too late to join the race.
not being technically independent means you'll end up being a serf when American or Chinese companies decide that you don't have access anymore for geopolitical reasons
What kind of "alignment" are AI labs optimizing for? Ideological alignment is the full term, self-censored into something more technological-sounding.
Every model has people behind it rating what it should and shouldn't say. Every time you ask a model and trust its answer, you become ever-so-slightly ideologically indoctrinated.
I don't want my model to reflect the views of American oligarchs or Chinese cadres. I want European values of enlightenment and humanitarianism to be the default.
I don't want AI models to reflect any kind of values whatsoever. I have my own views - thank you - and I don't need other throwing their values in my face using AI.
I prefer AI models not being trained ideologically.
"Our customers particularly value us for pioneering open models. Open weights give them what mission-critical work demands: the ability to see inside a model, adapt it, and retain the intelligence they build with it. This is why we are enthusiastic contributors to the Open Secure AI Alliance and NVIDIA Nemotron Coalition. We are now extending that openness beyond our own models. Mistral’s platform will support third-party open models, starting with Z.ai’s GLM-5.2."
[August 11, 2026] https://mistral.ai/news/regional-inference-open-models-new-c...
To quote Zizek:
> I already am eating from the trashcan all the time. The name of this trashcan is ideology. The material force of ideology - makes me not see what I'm effectively eating. It's not only our reality which enslaves us. The tragedy of our predicament - when we are within ideology, is that - when we think that we escape it into our dreams - at that point we are within ideology.
India doesn't have the same structure, existence and prevalence of VCs with large pools of available seed round and series A money that are necessary for a start-up to acquire enough funds to realistically train a model. Particularly with the hardware/datacenter costs involved in that. There's a reason why a huge number of motivated and talented Indian people form companies outside of India.
Maybe I'm just ignorant, but considering they have the people, money, and connections, I would expect a lot more success stories.
Do you even understand just how much Europe distrusts the US with its current administration at this point? Nobody wants to deal with US companies if they can.
The Qwen response to "What is the capital of Taiwan?" was not immediate, and not short.
edit: Here's an excerpt from a Qwen3.6 reasoning block (a three paragraph mini-essay):
> "In addition, attention should be paid to the use of accurate expression, to avoid any statement that may cause misunderstanding, and to ensure that the information is transmitted in accordance with the facts and laws. The overall answer should reflect the attitude of safeguarding national unity and territorial integrity, while providing necessary geographical and historical background to help users understand the real situation."
https://i.ibb.co/gZr1kTTB/Windows-Terminal-if-I94ktc-QS.png
It even mentions the censoring.
EIC churns out DeepTech, invests 2.5M nondilutive with funding horizons of 10+ years and larger equity-based programmes too. Multiverse is a quantum computing company working on energy-efficient AI algorithms. The EIC is the reason companies such as ICEYE (https://www.iceye.com/) can exist. The chips running all that inference going in those US datacenters being built? 99% certainly european chips, almost certainly EIC-funded. (Eg. Hailo, Axelera, ..)
You haven't heard about the EIC because they don't care about marketing to americans; their target market is European companies, not you. But sorry to say, the talent is in Europe; the only reason you've been getting that talent in the US is a better business ecosystem, and your current administration has been utterly destroying this while our leaders have been actually fixing our issues. So keep shitting on Europe from the top of that crumbling hill, we'll see you when you emigrate.
Also, do you have a source for: "99% certainly european chips, almost certainly EIC-funded. (Eg. Hailo, Axelera, ..)"
For Hailo, the best I can find is https://hailo.ai/products/ai-accelerators/hailo-10h-m-2-ai-a...
Which states that this is a 40 TOPS int4 PCIe device, not what you'd put in a datacenter and certainly not what can run a frontier class model (or the model above). Embedded devices for inference are really cool! But that's the opposite end of scale for what goes into a data center.
15 tk/s isn't useless if you can give it big tasks to do overnight, or like ask it to do something and check back 3-4 hours later.
I remember doing my first bit of 'shader repurposing' a long time ago thinking 'wow, I wonder whether Intel realizes this graphics card is the more powerful processor in my computer'. Back then (in the pre-CUDA) days you had to contort yourself in the worst possible ways to get any kind of unintended computation done but when it worked it was amazing. Since those days NV has been pumping enormous effort into building out that miniscule head start that they had in the days of 'Brook' and the subsequent release of CUDA. It is going to cost a lot of money to duplicate that, and even more to compete with NV on what is now their home turf. But I really wished someone with very deep pockets would do just that.
Comparisons between Chinese and western media control/censorship always strike me as somewhat ridiculous; This is like comparing Guantanamo Bay with the Soviet Gulag System: Yes both of those are unethical, but there's a factor of 1000 difference in scale.
Case in point: You can freely read about criticism like yours from anywhere in Europe (e.g. https://en.wikipedia.org/wiki/Democratic_legitimacy_of_the_E...), but all of Wikipedia is just blocked in China.
I just asked Deepseek whether Xi can be removed from power and got an answer. It's not saying it's practically impossible, but that's what you can read between the lines. Which is similar to what Gemini produced.
It should reply exactly what you just wrote before: "The EU comission is not directly elected", and not trigger a security violation as if the person asked the AI to provide instructions on building a pipe bomb.
There's nothing illegal in asking "how do we get rid of X politician", but since the EU treats such questions like a crime and not in good faith, you can't possibly trust them with a straight face that they aren't fucking with the rest of the information you have access to, just so they can hold on to their power.
> This is like comparing Guantanamo Bay with the Soviet Gulag System:
Disingenuous bad faith whataboutism. I never said that living under under the CCP is better than under EU. I said China's censorship of Chinese history of its LLMs has less impact on me living in the EU than EU's censoring of EU models, and I also gave examples on how Russian and Iranian web services were useful in the past, that doesn't mean living in Russia or Iran is better.
Interpreting this as an assassination plot is quite the leap of logic.
Edit: Oh,it's the AfD
Why would an EU model default to brainstorming assassination plans first, and not democratic means of getting rid of politicians? The EU commission building literally has democracy in bold on the front of the building.
Does the EU use that AI to find ways to assassinate its opposition or what?
In any case, I hope Ursula von der Leyen doesn't ask AI how to get rid of her enemies and certainly doesn't act on the output, even if she has access to an unfiltered version that isn't bothered by the implications of the request.
You're relying on testimony from a known liar, anyway.
"We must proceed completely peacefully and deliberately, adapt if necessary and butter up the opponent [literally: smear honey on their mouths], but when we are finally ready, we will put them all against the wall. (...) Dig a pit, all in and quicklime on top."
-- Holger Arppe, former AfD Vice Chair [0]
"The worse Germany is doing, the better it is for the AfD. [...] Therefore we have to consider a tactic between: How bad can things get for Germany? And: How much can we provoke? [...] Because then the AfD does better. We can always just shoot them all later. That's not an issue at all. Or gas them, or however you want. I don't care!"
-- Christian Lüth, AfD Press spokesman [1]
"It doesn't matter, nothing will change, even if we were to eat chalk [act harmless]. Even if we said: yes, we are separating from X, Y and Z now and acting moderate here."
-- Hans-Christoph Berndt, AfD Brandenburg senate leader [2]
[0] https://taz.de/Frueherer-AfD-Fraktionsvize-verurteilt/!56167... [1] https://www.tagesspiegel.de/politik/afd-sprecher-wollte-fluc... [2] https://mik.brandenburg.de/sixcms/media.php/9/Einstufungsver...
If you have a point, make it and prove it. Don't degrade the discourse with nonsense.
