https://techcrunch.com/2026/08/24/hugging-face-reportedly-in...
Nvidia agrees to acquire Hugging Face for $13B (businessinsider.com)
597 pointsby mfiguiere4 hours ago253 comments
https://www.theinformation.com/articles/nvidia-agrees-buy-op... (paywalled)
GeertB 3 hours ago
Nvidia's been pretty terrible for open source / free software. No need to quote Linus Torvalds here. They want to control what runs on their hardware. They want to you write code against their proprietary drivers and APIs, not directly against the hardware (which these days of course also contains plenty of software, but still).
Don't expect things to go differently this time around. Nvidia wants control over the software stack. Acquiring HF fits in perfectly. The play is long term.
rvz 2 hours ago
Nvidia has hardly open sourced the nvcc compiler, yet no-one here is complaining about why it is closed source.
Modular on the other hand creates the Mojo compiler gets criticised for not open sourcing it immediately and now once they do, no-one cares anymore.
Huggingface was not just a target for open source, but as a force to have open weight models run better on Nvidia against the rest.
silisili 2 hours ago
I assume it's because nvidia is a hardware company that produces world class hardware, and it's a compiler to target that.
Whereas mojo is a general purpose language, and we're absolutely spoiled for choice on modern languages with open source compilers.
I'm not saying it's fair or right, I still think mojo is neat, but isn't exactly comparing apples to apples.
rvz an hour ago
We can stop with these weak excuses since AMD and Intel have done more for open source than Nvidia has, including their GPU drivers for Linux.
Nvidia on the other hand has not and the best they have done is a bunch of closed-source blobs which they do more closed source releases than the rest.
Mojo is open source and targets all GPU architectures for their compiler regardless of the vendor and nvcc targets their own (and both that and CUDA are closed source).
So this is directly an apples to apples comparison.
latchkey an hour ago
> no-one cares anymore.
possibly because it took qualcomm buying them to make that happen.
rvz an hour ago
Just say you don't know.
Mojo was partially open source before Qualcomm bought them, and they were going to do open source it anyway.
Was NVCC or CUDA ever open source since the lifetime of its development?
latchkey an hour ago
> they were going to do open source it anyway.
uh huh.
> Was NVCC or CUDA ever open source since the lifetime of its development?
you ever ask Nvidia why? i did.
chii 25 minutes ago
> as a force to have open weight models run better on Nvidia against the rest.
this is the crux - if nvidia makes it so that open weights end up running better on nvidia hardware than competitor's, then it's going to prevent hardware innovation and competitiveness in the entire sector.
It's like as tho General Motors buys out oil refinery to make gas for all, but the gas somehow runs smoother in GM cars.
392 2 hours ago
I think it's a plausible play that keeps the AI boom in session for a year or two longer.
Or at least, $13b to stay at the head of the race (or keep the race running) must be worth it to someone's desk.
There's only $50b in datacenter buildout nationally (Source: Gemini, 2026).
So it is a bit of a puzzling choice for what amounts to a pile of software, in my opinion. but I don't know shit.
bbor 2 hours ago
They’re buying a brand, some employees, and some momentum — not any of their software. HF probably does have propriety goodies to make it all run efficiently, but certainly not a billion dollars worth, much less 13!
r0b05 2 hours ago
We are so used to these numbers being thrown around in the AI era that something one needs a reminder that this is 13 billion, not million. Insane exit by HF.
techpression 11 minutes ago
Was looking for this comment, the tech world has gotten completely insane with ”valuations”. I would love to hear why it was 13 and not 5. It would still be completely insane at 5 billion, but someone though they should add another 8…
wilburx3 2 hours ago
Yeah, the price is weird as fuck.
rezonant an hour ago
Did you just quote an LLM as a citation? That's not how citations work.
mapontosevenths an hour ago
"That's not how citations work." (Dude on the interwebz, 2026)
But more seriously, this is my first time seeing that as well, and I'm not sure I like it. Citing an LLM is a little like citing Wikipedia to me, you cite the primary source the LLM is quoting directly, not the secondary source.
lokar an hour ago
A Wikipedia link can be nice when you get several different sources provided.
Forgeties79 an hour ago
It’s the formatting and formality that bother me. Say “I looked it up with Gemini,” don’t give me some weak attempt at “proper citation” to legitimize the bare minimum effort you put in.
hobo_in_library 32 minutes ago
Why waste keystrokes when less do trick?
bdangubic an hour ago
2026 equivalent of “my Mama told me”
WarmWash an hour ago
Even worse is that Gemini's knowledge cutoff is still Jan '25.
zenoprax 20 minutes ago
> There's only $50b in datacenter buildout nationally (Source: Gemini, 2026).
Your reference lacks authority, veracity, and reproducibility.
dannyw 2 hours ago
NVIDIA is one of the most open labs.
They even share many of their pre-training and even post-training datasets for Nemotron on HuggingFace; for example: https://huggingface.co/datasets/nvidia/Nemotron-Post-Trainin...
Which other lab shares this?
Yes, there is no question NVIDIA wants to lock you into CUDA and their hardware. But also, they’ve consistently demonstrated the most openness when it comes to model training, datasets, and research; even before the LLM era (e.g. StyleGAN).
There’s also modelscope.cn (china’s huggingface) which is worth checking out. I would not be surprised if one day, we have to use China VPNs to download open weight models.
jrflowers 2 hours ago
lol “Everything you said is correct but you’re wrong because I’m thinking about something other than what you were talking about”
willy_k 2 hours ago
NVIDIA is a lot more than a “lab”.
matheusmoreira 2 hours ago
> NVIDIA is one of the most open labs.
Of course they are. They're commoditizing their complement.
I want to own the hardware, not play around in an nvidia fiefdom full of nvidia rules.
bdangubic an hour ago
nothing stopping you from doing that
Forgeties79 an hour ago
Are you joking? They just bought HF for $13bill and didn’t break a sweat. They are dominant and spreading.
behringer an hour ago
Can't buy a decent modern Intel gpu to run AI on because of nividias "donation"
cyberrock an hour ago
>No need to quote Linus Torvalds here
Evidently we should, because Linus has been more positive about Nvidia in the last 2 years [0]. I've been using the open driver for years now, for both gaming and CUDA.
[0] https://binarymusings.org/posts/talks/linus-on-ai-linux-in-k...
troyvit an hour ago
He definitely sounds more pragmatic than before, which is, in a sense, more positive.
> This is actually one of the benefits brought by AI; it has made Nvidia a good participant in the Linux kernel space. [...] Now, when Linux is so important for AI clouds, Nvidia suddenly cares very much about Linux.
jijji an hour ago
I for one welcome our new robot overlords.
lambda 7 minutes ago
Nvidia releases some of the most open open weights models, Nemotron 3, which have the full training code open, and most but not all of the training datasets.
Nvidia is a big company. They are good about some things and bad about others.
I think they really do like open weights because they make some of the best hardware for training, and the more open weights models there are, the more people are training and fine-tuning them, mostly on Nvidia hardware.
I feel like Nvidia is one of the better choices for buying Huggingface. Not perfect, but definitely far from the worst.
manlymuppet 3 hours ago
Potentially horrible for monopoly reasons, but if other big acquisitions in the AI era teach us anything, developers are about to get a whole lot of free and discounted trial credits.
That’s at least a plus. I will happily burn through as much VC money as they will give me to tinker with my projects.
abixb 3 hours ago
Yes. Great for devs, for a while. Awful for overall competition and industry's health.
I think the federal antitrust regulators are asleep.
Edit: antitrust regulators' job has just begun -- we'll see how this deal gets adjudicated by the FTC (if at all).
herpdyderp 3 hours ago
Except for during Biden, they've been asleep for almost 50 years.
abixb 3 hours ago
Lina Khan @ the FTC was the best thing about the Biden administration, for me.
timcobb 3 hours ago
I miss the Biden administration
jacquesm 3 hours ago
Large numbers of people outside of the United States probably do.
Gigachad 2 hours ago
China and Russia are loving things right now.
nozzlegear 2 hours ago
I miss him from inside the United States too.
platevoltage 2 hours ago
That's a sad thing to have to say, but yeah.
Danox an hour ago
I miss every administration back to Roosevelt.
smelendez 3 hours ago
I think failing to take advantage of the charisma of both Lina Khan and Janet Yellen, and not making them public-facing economics gurus like Alan Greenspan or Larry Summers historically were, was a huge tactical mistake by Biden.
tiahura 2 hours ago
Yellen was objectively the dumbest Treasury Secretary ever. When everyone and their dog realized inflation was getting out of control, she thought it was transitory. She literally gets an F- as an economist.
And some former Roomba employees may have thoughts about Kahn.
Grombobulous 2 hours ago
Kahn was right about Roomba. Amazon was going to torpedo all the superior competition on the Amazon platform where almost all robot vacuums are sold in the US, plus as the e-commerce monopoly, Amazon was going to instantly have access to an army of cameras inside people’s homes to analyze personal information related to purchase preferences.
IMO allowing Roomba to hit chapter 11 was still a better option for the consumer than handing them to Amazon. They’re still in business as an independent competitor on the market, consolidation was successfully avoided.
Yellen, I don’t have much of an opinion on, but she absolutely wasn’t alone in that opinion (the nuance of that opinion being inflated by this hyperbole), and the treasury is a lot less involved than the federal reserve in doing anything about inflation, anyway.
kaonwarb 2 hours ago
iRobot is now 100% owned by Shenzhen Picea Robotics of China. (https://www.sec.gov/Archives/edgar/data/1159167/000115916726...)
Grombobulous an hour ago
Yes, that's a very good company to be the owner of iRobot, because they are just a robot factory and not a near-monopoly e-commerce retailer.
They are essentially on equal footing with other robotic vacuum manufacturers. iRobot didn't go out of business or get absorbed into a larger company lowering competition in the marketplace.
It's also not the FTC's job to ensure that companies, especially ones with zero/trivial national security or domestic labor force value, remain under domestic ownership. Amazon itself is not really a "domestic" company, it's publicly traded. Anyone from any non-sanctioned country can buy shares of Amazon.
Grombobulous an hour ago
This is a tangent to my first comment parent to this one, but also more related to the actual article at hand: whether it's Amazon buying Roomba or Nvidia buying Hugging Face, I have a somewhat radical (or is it?) opinion that large companies like Nvidia, Amazon, Apple, Microsoft, Coca-Cola, etc, should not actually be allowed to acquire companies. At all. under any circumstances, even if competition is healthy.
These companies are generally large enough that they do not need the competitive aid of buying an existing company and starting with that sort of structural advantage.
E.g., did Nvidia not have enough money in their bank to start a company to compete with Hugging Face? This is a company that reportedly has ~200-300 employees with investment rounds totaling $400 million. Nvidia made $31.9 billion in net income last quarter.
I look at a company like Xiaomi which just developed its automotive division in-house without resorting to buying car companies. I think our traditional business and finance mindset has an overreliance on acquisitions.
creato an hour ago
Roomba was not competitive in the market, no amount of Amazon market manipulation would have changed that. Best case scenario, Amazon would have invested in making it competitive. Worst case, their engineers would have been absorbed into Amazon's warehouse robotics projects.
Is either of those worse than what actually happened: the company is now a zombie brand for a Chinese company?
Grombobulous an hour ago
Of course Amazon would have changed that. You would go on Amazon and search for "robot vacuum" and Amazon would put iRobot at the top of the results. Review manipulation on the platform would be trivial.
Amazon could email/push notification/text customers asking for reviews of iRobot vacuums but then not do the same for competing vacuums, skewing their reviews higher (asking for reviews boosts ratings by gathering opinions from happy customers who usually don't bother writing a review). Competing brands potentially don't even have your contact information to ask for a review.
Go on Amazon right now and search for "usb cable." There's a giant banner at the top that recommends the Amazon Basics brand, three across horizontally, which on my desktop monitor takes up nearly 50% of the screen real estate. Then below it are the Anker cables that you're more likely to be looking for.
They would have almost certainly been manipulating pricing on them as well. For example, they could take the strategy of lowering the price of the vacuums to break even or sell as a loss leader, but use them as a data-gathering robot in your house to help Amazon sell more of everything else. They could make their Alexa smart home platform preferential to iRobot or lock out competing models.
Robot vacuums are a consumer goods category that is heavily skewed toward Amazon.com as the place of purchase compared to other retailers.
I think "zombie brand owned by a Chinese company" is actually preferable, yes. They still operate and sell vacuums competing with the other robot vacuums on the market, and they aren't in service as household data collection bots for a monopoly e-commerce platform.
I don't think the ownership of the company being foreign or domestic is very relevant to the FTC's goal of preserving positive trade conditions. Would we think the iRobot situation was a bad outcome if iRobot was purchased by a foreign company we view more positively like Miele? We only think of it negatively due to anti-Chinese bias. iRobot being Chinese-owned is almost certainly the best possible outcome for preserving the amount of competition in the market.
qchris 2 hours ago
The company's name was iRobot, not Roomba.
I spent a little time working there around 2017. Whatever you're implying my former colleagues' thoughts on the subject are (or mine, for that matter), you're probably wrong.
theendisney 2 hours ago
Now that you mention sleeping. I imagine one could do cool sleeping startups that clone something but exist only on paper. The sleeper startup is only launched the moment some ham fisted mega corp aquires the original and all the customers are looking for a way to abandon ship.
alightsoul 2 hours ago
You can torrent and seed these models via torrenting. They are not piracy so there's no problem hosting them besides maybe not having enough bandwidth
matheusmoreira an hour ago
> I think the federal antitrust regulators are asleep.
Have been for a long time. All of the big techs should have been broken up long ago.
rablackburn 3 hours ago
> That’s at least a plus. I will happily burn through as much VC money as they will give me to tinker with my projects.
The era of VC-funded home-delivery recipe boxes is still my favourite.
Eji1700 3 hours ago
Yeah this is the "early netflix" era. Take advantage while you can and plan for the future.
Gigachad 2 hours ago
Counter point is China can afford to keep pumping out free models for long enough to crash the AI industry, As well as spin up memory production.
Its likely in the near future we will be able to buy 128gb mac minis and run local AI for free.
bbor 2 hours ago
If we survive, that is.
nickysielicki 35 minutes ago
Why would China want to crash the AI industry? Their goal is not to damage the United States, necessarily. Their goal is to have the best domestic AI on the planet. They intend to do that through the way that they dominate every other industry: good enough quality at a much greater scale.
alightsoul 2 hours ago
That means we will torrent these ai models. They are not pirated so there's no problem doing so
root_axis 3 hours ago
True, but better nvidia than any other big tech contender. Nvidia has a very strong business incentive to make huggingface thrive, whereas pretty much everyone else has the opposite incentive.
typ 2 hours ago
Especially when all major labs are building their custom inference chips.
make3 2 hours ago
I think this is good.. They have incentives to keep things free to keep people using their GPUs. I think that's one of the least enshittifying outcomes possible
drTobiasFunke 35 minutes ago
Nvidia is a public company. What VC money?
binarymax 3 hours ago
Well congrats to Clem and the team. I remember when huggingface was doing things like coreference resolution models on spacy.
I hope nvidia does right by the community.
Edit to add: $13B should cover the S3 egress fees for a couple months :D
kfse 2 hours ago
The $13b doesn't go to hugging face's bank acct to pay for anything like S3 egress fees... It go to hugging face's owners' Bank accounts for them to do anything else. New cohorts of billionaires and centi millionaires getting minted.
heliosAtwork an hour ago
They need to switch to egrees free cloudflare R2!
rcleveng an hour ago
Totally, and as long as they don't turn on the AWS NAT Gateway, they'll be just fine.
esjeon 3 hours ago
Obviously, NVIDIA is trying to own the AI development chain.
Owning HF -- the discovery and distribution channel -- is one thing, but I think the biggest threat vector is the privileged access to HF platform data, that includes HW survey info and model download pattern. This can be a borderline anti-trust case.
Fordec 39 minutes ago
If the acquisitions of late tell us anything at all, it's that Nvidia wants to be a vertical AI company, not a hardware or gaming company.
Conol_ai 2 hours ago
The more interesting is HF turned down a $500M Nvidia investment late last year at
a $7B valuation, after passing on a $235M round in 2023 at $4.5B —
going from "we don't want a dominant investor" to a $13B full
acquisition in under a year is quite the reversal.
transitorykris 2 hours ago
Well, if they didn't want a dominant investor, they didn't get one, they got acquired.
alansaber 2 hours ago
Sounds like it was just too much money for the stakeholders to refuse
dingdongditchme an hour ago
Yea, my bet is that there is quite a long list of money bags sitting behind the leadership team that will weigh in if the offer is juicy enough.
moralestapia an hour ago
Seems the opposite of a reversal to me.
Reversal would be "valued at 7B then valued at nothing", this is more like "nah, we want more" then "nah, we want more" then "yes, that's what we want :)".
altmanaltman an hour ago
I don't think they passed on the 4.5 billion. They still raised it in 2023 with investors like Alphabet etc. They turned Nvidia *earlier this year* for the $7 billion valuation deal. It makes sense if they were angling for an acquisition.
This is the article you are referring to is this but your facts are wrong: https://techcrunch.com/2026/08/24/hugging-face-reportedly-in...
kpw94 3 hours ago
Remember just 6 months ago that "Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI"
(https://news.ycombinator.com/item?id=47088037)
(Ggml.ai is llama.cpp.)
Curious if the “I consider HuggingFace more "Open AI" than OpenAI” sentiment in that top comment will still apply with NVIDIA as the boss now...
make3 2 hours ago
NVidia has reasonable incentives to keep things open indefinitely, it wants people to use it's GPUs
noosphr 2 hours ago
Issue is that llama.cpp is the best way to run models on hardware that isn't nvidias.
alightsoul 2 hours ago
Except when they have less than 16 gb of ram?
swozey an hour ago
It'd be really nice if I could use my egpu 4090 on my macbook pro..
heliosAtwork 2 hours ago
Nvidia is promoting "unmetered intelligence". Jensen wants everyone to have a spark dgx with an nvidia chip in their home.
petterroea 3 hours ago
I have never seen an acquisition be good for the users. I struggle to believe this one will be different
bulbar 3 hours ago
Acquisition of GitHub let to increased downtimes, but also to unlimited private repos for the free tier.
unsungNovelty 2 hours ago
... Which gitlab already had before github. And still does.
petterroea 2 hours ago
The unlimited private repos for free tier would have happened anyways. You have to remember that this was around the time people were jumping ship from GitHub exactly because they didn't offer private repos for free. It was something many wanted, and they were actively seeking out other places that would give it to them.
akulbe 2 hours ago
Truer words were never spoken. This is painfully true.
alansaber 2 hours ago
The fit is theoretically good (nvidia is model agnostic) but I can see then pushing the ecosystem increasingly towards cuda and some custom nvidia software packages for inference etc etc
alightsoul 2 hours ago
Yeah I'm afraid they will only keep cuda support for their software. Good thing it's all Open source
blihp 2 hours ago
I'm sure it'll be fine. As long as you have the latest 90-class card you'll probably continue have nearly full functionality from the HF libraries. (Upgrade to enterprise gear for full functionality)
1123581321 an hour ago
Apple buying Workflow (became Shortcuts). Its users have a better, faster tool and got it soon after acquisition. And it has many more users now.
sixothree an hour ago
Was Workflow available on every platform running on every hardware provider?
1123581321 an hour ago
I’m not sure your point, sorry.
Gigachad 3 hours ago
What is the business model of hugging face? My understanding was they were basically just a file hosting platform.
dylan604 3 hours ago
Isn't this pretty much the model for any tech company? Rush to get MVP released, stoke the hype, cash in on FOMO.
RachelF 3 hours ago
I guess we may have to start paying to download models.
Or perhaps they will start throttling downloads for free users.
I don't know what they business case is, it might be to shut them down: I suspect good free models on local hardware is a threat to Nvidia's investments in OpenAI/Anthropic.
esseph 3 hours ago
Maybe they're trying to see how big the market actually is before considering shutting it down, maybe or getting lawyers and politicians to try and outlaw or restrict open models if they see a big enough opportunity.
sebmellen 3 hours ago
I see the opposite — NVIDIA is aiming to commoditize the model layer and push open models that are tuned to run on NVIDIA hardware.
a_wild_dandan 3 hours ago
Maybe everyone will migrate to Hugging Bay for downloading models via torrent?
Gigachad 3 hours ago
File hosting isn’t that hard to replicate though. Even if they start charging people will just set up torrents for the files or lists of mirrors.
Unless I’m missing something, this feels like Nvidia having more money than they know what to do with.
dragonwriter 3 hours ago
NVidia has been expending energy helping improve local models and inference platforms for them targeting NVidia GPUs; good free models that users can run locally rewards Nvidia’s investment in product lines for local inference (DGX, RTX PCs, etc), as well as—given their continued dominance in the space—the premium over competitors of their consumer and workstation GPUs.
dragonwriter 3 hours ago
They have various paid services that relate to AI—paid inference hosting; paid accounts aimed at AI development with GPU rentals [credits plus paid overages, I believe], more private storage and public storage than free accounts; and many of the community a and some other benefits on individual/team/enterprise tiers; additional paid storage above the base quotas for the paid account tiers; on demand rentals of HF managed containers on GCP and AWS, and some other things.
andy99 3 hours ago
Didn’t we just witness it? Become a critical ecosystem player then get acquired based on the value of that position.
Gigachad 3 hours ago
I’m asking more to work out what is the basis of this valuation. Why would Nvidia spend $13B for what seems to be a services that gives things away for free.
weird-eye-issue 3 hours ago
This is like saying GitHub or Google just gives away everything for free
Gigachad 3 hours ago
Not really. GitHub has a strong platform of paying customers and a solid product that’s hard to replicate.
Google is obvious.
Hugging Face is a file download mirror with a couple of side features dangling off.
weird-eye-issue 2 hours ago
Whether you realize it or not Hugging Face's monetization is just as obvious as GitHub's
calebkaiser 2 hours ago
They hit 150 million in annual recurring revenue this year.
Pretty nice dangling side features apparently.
Gigachad 2 hours ago
That still leaves the company valued at 86 years of revenue
calebkaiser 2 hours ago
Yeah, great point. Nvidia could potentially be overpaying. Not sure how that equates to Huggingface being "a file download mirror with a couple of side features dangling off".
weird-eye-issue an hour ago
Just wait until you hear that Facebook bought Instagram for $1 billion in 2012, when it was generating no revenue.
Gigachad 21 minutes ago
While it's probably easier to say this in retrospect, they were eliminating a direct competitor to their core business. Something so advantageous it should have been blocked by regulators.
I can't see this as being as good of a purchase, especially when it's 13x the price of what was seen as an absurdly large amount back then.
weird-eye-issue 16 minutes ago
You are very good at bringing up irrelevant numbers and missing the forest for the trees
Mistletoe 43 minutes ago
Absolutely none of this makes sense and the thing that amazes me is how long it has continued. Future historians will just laugh at how stupid and obvious the crash was.
jdejean 3 hours ago
NortySpock 3 hours ago
Risk management.
Nvidia has a market cap of $5T USD today, and a decent chunk of that is due to LLM speculation.
Does Nvidia want their stock price to be at risk of being tanked by a download service being in the news? No, they want to make sure the party keeps going and is under their direct supervision, and part of that is making sure Hugging Face isn't bought by a competitor or runs out of money.
DrewADesign 3 hours ago
Talking about profit is so passé in the new economic paradigm. The rules have changed— it’s about how much a company is worth. Get with the times.
–Some guys in every bubble I’ve witnessed.
lotsofpulp 2 hours ago
This gets mentioned a lot, but the most valuable companies, by orders of magnitude, have the most profit, by orders of magnitude.
calebkaiser 2 hours ago
I dunno. You can argue over whether they're overpaying, but it's not like Huggingface is Clinkle. They hit $150 million in ARR this year, they have tons of runway, and according to reports, have just started to even burn the money they raised a few years ago.
I get that it's fun to be glib about the stupidity of tech elites and investors in general, but Huggingface have been pretty open about their financials and are, in my opinion as a practitioner in the field, one of the most responsible orgs in our space. They've been a pillar of open source ML for years now and have made a very positive impact on our ecosystem.
Nvidia is getting a real business generating revenue, and the center of the universe for open models. Both seem like pretty valuable attributes, from Nvidia's perspective.
jameshart 43 minutes ago
150 million a year? Why, at that rate NVidia will make their money back in just 86 years.
calebkaiser 15 minutes ago
Solid point. I'm sure Nvidia went into this deal expecting completely flat growth and no other benefits to their core business. Sorta like how Meta never increased Instagram's revenue from $0 and is still waiting for it to pay off that billion dollar acquisition price.
Or like GitHub, which was generating something like 200 million in ARR and had never hit profitability when Microsoft bought it for $7.5 billion back in 2018. I'm sure it has come as nothing but a happy surprise to Microsoft that GitHub generated $1 billion in 2023. They had initially penciled it in for 38 years til ROI.
vgeek 2 hours ago
I think Russ Hanneman calls it a "pure play" type of company. He put radio on the internet, so he would know.
zx8080 2 hours ago
> Why would Nvidia spend $13B for what seems to be a services that gives things away for free.
To stop that! Literally. To stop those services being free.
a34729t 2 hours ago
for 6 months?
strictnein an hour ago
Why would Nvidia destroy the first place we all look for models to load on our Nvidia hardware?
ks2048 2 hours ago
They make it easy to run models. The models have to run on hardware somewhere.
strictnein an hour ago
Possibly, but they were profitable already.
itemize123 3 hours ago
ondemand cloud? b2b?
raincole 2 hours ago
If GLM-5.3 was truly trained and ran on Chinese chips, we'll have to ask what the business model of Nvidia itself is soon :)
0xbadcafebee 2 hours ago
it's literally the GitHub for AI
Gigachad 20 minutes ago
Github was acquired for $7B. It's hard for me to see how HF is worth almost double Github.
andy99 3 hours ago
The open weights AI ecosystem is way too concentrated with HF. That said, it was clear they were going to either get bought or aggressively monetized at some point, Nvidia seems like it could be one of the more aligned acquirers.
I do worry about any sort of crowding out or downplaying non Nvidia-relevant quants, and about changing rules to crack down on models or datasets that for one reason or another “don’t align with their corporate values” - uncensored etc. Someone mentioned Microsoft and GitHub, they appear to me anyway to have been very hands off, I hope it’s the same model.
zb3 3 hours ago
We also have modelscope.cn, I'm sure that one won't get acquired
alightsoul 2 hours ago
It's already part of Alibaba, which has its own incentives. So does kaggle with Google. Civitai is neutral but it is mostly nsfw even though they don't show it outright. Torrenting is the Best way
jacquesm 3 hours ago
There are always torrents, which would seem to be a fine way to distribute models.
brainless 3 hours ago
HF is the default platform to find models. Not just ones published by big labs but also a lot of the distilled or fine-tuned versions, etc.
If Nvidia buying HF makes it tough for all the diverse models on HF, then what are some alternatives?
It seems models are the best things to be available on a Torrent platform? Of course HF is much more than just the files but perhaps the metadata can be separate and hosted on multiple community platforms.
ekianjo 2 hours ago
> what are some alternatives
good old torrents
pwython 31 minutes ago
There's over 3 million models on HuggingFace. So many quants. I'd imagine less than 0.5% would actually be seeded.
chorizo 2 hours ago
If HF didn’t exist, model releases would likely have been on torrents like Linux iso’s back in the day.
swozey an hour ago
I'd honestly prefer that. I'm usually on 5g if not 4g and pulling models is an absolute nightmare when I basically have to use HF and even with a token getting throttled. Although I have a pretty unique use case. Pulling wads of 5gb tensor files over cellular is painful.
brainless 34 minutes ago
I am on the same boat, I live in a small Himalayan village. I have 2x5G based devices and one Wireless bridge (Ubiquiti LiteBeam M5) for a local broadband. Generally I get 50 Mbps, sometimes up to 100 Mbps
Fervicus 2 hours ago
This is the world we live in now. 3-4 megacorps owning everything in every major product category - eyewear, cosmetic, consumer goods, media, tech; you name it.
alightsoul 2 hours ago
We need small companies worldwide, developing the things huggingface does. The us china and Europe dominate while the rest of the world sits by idly. Torrent and seed ai models. It's not piracy
unglaublich 21 minutes ago
If you can't beat them, buy them. Hallmark of late-stage capitalism.
numberwan9 2 hours ago
I really hope they retain and pay Georgi Gerganov. We are toast without ggml.
peri-cl 2 hours ago
I remember when people realized gaming cards could be used for general compute (GPGPU) on certain types of parallel numeric work. It was a very cool thing for many types of hackers and science and engineering students and NVIDIA went far out of their way to ensure we could never buy it.
The optimal market strategy there (as in a lot of places) wasn't "sell as much as you can". There's often a superior strategy, when (as with HPC) you have minority industry customers who are very rich and have low price sensitivity. It's to raise the price to what those special customers are willing to pay, and to drop everyone else.
What NVIDIA did was to rip out FP64 capability, systematically, from all of their consumer cards. They firewalled off "useful for GPGPU" as a differentiating feature, segmented the market, and astronomically raised the price of what (if you were looking soley at cost-to-manufacture) could have been easily affordable to any ramen student.
(It's a more obscure version of the Intel-made-ECC-memory-disappear story).
See, e.g.
https://news.ycombinator.com/item?id=47068890 ("15 years of FP64 segmentation, and why the Blackwell Ultra breaks the pattern (nicolasdickenmann.com)")
RachelF 2 hours ago
Yes, this is ironic - it was the tinkerers using CUDA on cheap graphics cards that made Nvidia stuff useful in constrained academic environments.
True about Intel and ECC, but AMD now does similar things, even with their consumer CPUs and chipsets.
These three companies now make very sure that consumer products can never canibalize those juicy data center profits - so they make sure to limit what the consumer segment can do.
vmg12 3 hours ago
The companies have aligned incentives so I don't think this is a terrible acquisition for customers.
rileymat2 2 hours ago
Vertically integrated monopolies naturally have aligned in incentives for themselves, which allows more capture which turns out bad for customers. Whether it is a monopoly is an open question.
vmg12 an hour ago
This isn't vertical integration. This is a commoditize your complements play and that does benefit customers. Nvidia benefits from supporting the open model ecosystem, if models are a commodity the gpus become what's scarce.
etempleton 2 hours ago
I am just curious, and I genuinely want to know, how is a 150 million a year in revenue company worth 13 billion dollars?
rvz 2 hours ago
Same reason why GitHub was acquired for $7.5B at $250 million ARR which is 30x revenues:
VCs could not see any other reason to raise more money and Huggingface was not growing as fast as they thought to justify the valuation or the next fundraise.
So they might as well get Nvidia to save them from the VCs pressurizing them.
Jonathanvw 2 hours ago
That doesn't make much sense to me. Companies don't just buy companies for absurd multiples out of the niceness of their heart.
In the case of GitHub, it was likely for data reasons + wanting to own where developers do work (VScode + Github).
In the case of HuggingFace, honestly not sure as I'm not familiar enough with their business. But I can assure you that Nvidia didn't buy them for 13 billion cause HuggingFace were desperate. When you're desperate, you sell for less not more.
iamflimflam1 2 hours ago
Somewhere, there’s a slide deck that shows future revenue going hockey stick.
Joel_Mckay 23 minutes ago
Curvature and convexity matters even in "hockey-stick" models =3
bensyverson 3 hours ago
I hope they're good stewards. This reminds me of the Microsoft GitHub acquisition in terms of the importance to the broader community.
impulser_ 3 hours ago
I think they will be, Nvidia has released open source model that also included the data it was trained on. I think they are the only ones that have done that.
Vespasian 3 hours ago
Handing out digging spots if you are the shovel manufacturer sounds reasonable.
bensyverson 3 hours ago
As long as the spots continue to be compatible with shovels from other shovel manufacturers
esalman 3 hours ago
As someone who've played video games and obsessed over graphics cards long before this all started, don't get your hopes up.
diabllicseagull 3 hours ago
as gamers, we got burnt by nvidia way too many times. I'm still holding grudge over shutting down 3dfx.
isatty 3 hours ago
Nvidia is indeed a terrible company when it comes to open source, etc. however, as someone who has been gaming on desktop hardware since the 6xxx card days, Nvidia hardware has always worked. I switched to Linux full time since the 2XX days (maybe it was the 280 gtx?) and never had a lot of the problems everyone else got on Linux (using only the closed source drivers).
jacquesm 3 hours ago
Likewise, their hardware has been rock solid and very performant for me. I bought most of my current setup second hand and they've been going strong for years now.
esalman an hour ago
NVIDIA introduced programs like The Way It's Meant to be Played, optimizing major game titles specifically for NVIDIA architecture, which made competitions like AMD cards underperform in critical releases.
I'm old enough to remember.
TiredOfLife an hour ago
Amd and Intel have the same programs.
esalman 37 minutes ago
But they don't/didn't have the same level of market share as Nvidia, particularly in the PC gaming market. Nvidia had 65%-85% market share, and still worked with game studios to make sure their competitions failed. That's the opposite of good stewardship that's all I'm saying.
isatty 19 minutes ago
You are being disingenuous.
They worked with studios to optimize games for their hardware. I don’t think they went around sabotaging other companies hardware. They did what was good for them.
Why did competition not do the same?
isatty 36 minutes ago
I fail to see how this is a problem?
AMD could’ve done the same.
TiredOfLife an hour ago
3dfx was going through bankruptcy
jazzpush2 an hour ago
NVIDIA had more open-source models on huggingface than any other company. A lot better them than any other frontier lab, tbh.
dnnehgf an hour ago
stripe buys openrouter. nvidia buys hugging face. everyone on the hunt for the app that will function as hub/broker/search/marketplace for models.
austinbaggio an hour ago
Monumental implications on OS models. Proper investment in US OS incoming? HF has the best possible distribution to that audience
janalsncm 3 hours ago
They were already pretty vocal about it, but this makes Nvidia pretty much the de facto face of open weight models.
As battle lines get drawn over duopoly vs. open weight it’ll be interesting to see what Nvidia does. They definitely want a piece of more of the stack especially as Huawei chips become more and more of an alternative to cuda.
alightsoul 2 hours ago
Alibaba already has modelscope and their own chips, and qwen models. Nvidia is doing the same with chips, huggingface and nemotron models
akulbe 2 hours ago
Like some other folks, I wonder whether this will be a good thing, or a bad thing.
I feel like if Nvidia ends up turning into a bad actor, in terms of restricting/censoring models... another HF will spring up.
Danox an hour ago
Not good news and in the long run this circular move will fail for them. What are the alternatives?
aizk 3 hours ago
Fix your title it's not finalized yet!
girfan 3 hours ago
According to The Information, it seem it's finalized? https://www.theinformation.com/articles/nvidia-agrees-buy-op...
underdeserver 13 minutes ago
Then use that as the link.
Everything I can find online is referring to this one source. That doesn't tell me if it's happening or not.
cmiles8 22 minutes ago
Congrats to the team. I have a strong feeling this will go down in history as one of the last great exists before the bubble bursts.
12-24 months from this acquisition will likely look like a crazy burn of capital and cash.
__mharrison__ 3 hours ago
The lowest hanging fruit that hf could offer is community monitored commands for running local models.
Right now it is a pain to find the correct incantation.