I think this is a game changer, because data privacy is a legitimate concern for many enterprise users.
Btw, you can also run Mistral locally within the Docker model runner on a Mac.
Not quite following. It seems to talk about features common associated with local servers but then ends with available on gcp
Is this an API point? A model enterprises deploy locally? A piece of software plus a local model?
There is so much corporate synergy speak there I can’t tell what they’re selling
This announcement accompanies the new and proprietary Mistral Medium 3, being discussed at https://news.ycombinator.com/item?id=43915995
While I am rooting for Mistral, having access to a diverse set of models is the killer app IMHO. Sometimes you want to code. Sometimes you want to write. Not all models are made equal.
Tbh I think the one general model approach is winning. People don't want to figure out which model is better at what unless its for a very specific task.
Couldn't you could place a very light weight model in front to figure out which model to use?
Well that sounds right up the alley of what I built here: www.labophase.com
Why use this instead of an open source model?
> our world-class AI engineering team offers support all the way through to value delivery.
Guess that makes sense. But I'm sure they charge good money for it and then you could just use that money for someone helping you with an open source model.
[dead]
I really love using le chat. I feel much more save giving information to them than to openai.
interesting take. i wonder if other LLM competitors would do the same.
I don't see any mention of hardware requirements for on prem. What GPUs? How many? Disk space?
I'm guessing it's flexible. Mistral makes small models capable of running on consumer hardware so they can probably scale up and down based on needs. And what is available from hosts.
GPT4All has been running locally for quite a while...
Another new model ( Medium 3) of Mistral is great too. Link: https://newscvg.com/r/yGbLTWqQ
Interesting. Europe is really putting up a fight for once. I'm into it.
Expected this comment.
Mistral has been consistently last place, or at least last place among ChatGPT, Claude, Llama, and Gemini/Gemma.
I know this because I had to use a permissive license for a side project and I was tortured by how miserably bad Mistral was, and how much better every other LLM was.
Need the best? ChatGPT
Need local stuff? Llama(maybe Gemma)
Need to do barely legal things that break most company's TOS? Mistral... although deepseek probably beats it in 2025.
For people outside Europe, we don't have patriotism for our LLMs, we just use the best. Mistral has barely any usecase.
> Need local stuff? Llama(maybe Gemma)
You probably want to replace Llama with Qwen in there. And Gemma is not even close.
> Mistral has been consistently last place, or at least last place among ChatGPT, Claude, Llama, and Gemini/Gemma.
Mistral held for a long time the position of "workhorse open-weights base model" and nothing precludes them from taking it again with some smart positioning.
They might not currently be leading a category, but as an outside observer I could see them (like Cohere) actively trying to find innovative business models to survive, reach PMF and keep the dream going, and I find that very laudable. I expect them to experiment a lot during this phase, and that probably means not doubling down on any particular niche until they find a strong signal.
>You probably want to replace Llama with Qwen in there. And Gemma is not even close.
Have you tried the latest, gemma3? I've been pretty impressed with it. Altho I do agree that qwen3 quickly overshadowed it, it seems too soon to dismiss it altogether. EG, the 3~4b and smaller versions of gemma seem to freak out way less frequently than similar param size qwen versions, tho I haven't been able to rule out quant and other factors in this just yet.
It's very difficult to fault anyone for not keeping up with the latest SOTA in this space. The fact we have several options that anyone can serviceably run, even on mobile, is just incredible.
Anyway, i agree that Mistral is worth keeping an eye on. They played a huge part in pushing the other players toward open weights and proving smaller models can have a place at the table. While I personally can't get that excited about a closed model, it's definitely nice to see they haven't tapped out.
It's probably subjective to your own use, but for me Gemma3 is not particularly usable (i.e. not competitive or delivering a particular value for me to make use of it).
Qwen 2.5 14B blows Gemma 27B out of the water for my use. Qwen 2.5 3B is also very competitive. The 3 series is even more interesting with the 0.6B model actually useful for basic tasks and not just a curiosity.
Where I find Qwen relatively lackluster is its complete lack of personality.
I certainly had some opposite experiences lately, where Mistral was outperforming Chatgpt for some hard questions.
Whats your point here? There is a place for a European LLM, be it “patriotism” or data safety. And dont tell me the Chinese are not “patriotic” about their stuff. Everyone has a different approach. If Mistral fits the market, they will be successful.
[flagged]
That’s what it already is, bc that’s the raison d’etre for mistral: a small effort to arrest Europe’s long slide towards a branch-plant economy, a market for goods not a place to innovate, a continent which no longer remotely holds superpower status. Client state, client economy.
The EU enjoys making such performative attempts so they can maintain a facade of being a serious player with a workable future vision, in lieu of serious change (cut welfare to the bone, rebuild militaries, slash taxes and regs) that would actually help.
Mistral’s only two major advantages have been openness and european ness. Since openness is gone (understandably since it’s not super monetizable) european is basically all it has left.
You are probably getting downvoted because you don't give any model generations or versions ('ChatGPT') which makes this not very credible.
Its more likely that I'm getting downvoted by patriotic Europeans who came into a thread about an European company.
But ChatGPT has always been state of the art and cutting edge. Do I need to compare the first mistral models to 3.5? Or o4 and o3?
Does any reasonable person think Mistral has better models than OpenAI?
In your first comment you mentioned you used Mistral because of its permissive license (so guessing you used 7B, right?). Then you compare it to a bunch of cutting edge proprietary models.
Have you tried Mistral's newest and proprietary models? Or even their newest open model?
To me that's a clue that it's written by an American :-).
I love that "le chat" translates from French to English as "the cat".
Also, "ChatGPT" sounds like chat, j’ai pété ("cat, I farted")
Mistral should highlight more in their marketing that it doesn’t make you fart.
Instead it disobeys commands, uses up your resources then you find it never belonged to you in the first place.
Their M logo is a pixelated cat face as well.
I wonder if they mean to reference the Belgian comic Le Chat by Philippe Geluck.
This will make for some very good memes. And other good things, but memes included.
Mistral models though are not interesting as models. Context handling is weak, language is dry, coding mediocre; not sure why would anyone chose it over Chinese (Qwen, GLM, Deepseek) or American models (Gemma, Command A, Llama).
Command A is Canadian. Also mistral models are indeed interesting. They have a pretty unique vision model for OCR. They have interesting edge models. They have interesting rare language models.
And also another reason people might use a non-American model is that dependency on the US is a serious business risk these days. Not relevant if you are in the US but hugely relevant for the rest of us.
Data privacy is a thing - in Europe.
I flip back and forth with Claude and Le Chat and find them comparable. Le Chat always feels very quick and concise. That's just vibes not benchmarks.
Too little too late, I work in a large European investment bank and we're already using Anthropic's Claude via Gitlab Duo.
Is there are replacement for the Safe Harbor replacement?
Otherwise it could be illegal to transfer EU data to US companies
The law means don’t do what a slow moving regulator can and will prove in court. In this case, the law has no moral valence so I doubt anyone there would feel guilty breaking it. He may mean individuals are using ChatGPT unofficially even if prohibited nominally by management. Such is the case almost everywhere.
There are plenty of other ways to run Mistral models on a Mac. I'm a big fan of Mistral Small 3.1.
I've run that using both Ollama (easiest) and MLX. Here are the Ollama models: https://ollama.com/library/mistral-small3.1/tags - the 15GB one works fine.
For MLX https://huggingface.co/mlx-community/Mistral-Small-3.1-24B-I... and https://huggingface.co/mlx-community/Mistral-Small-3.1-24B-I... should work, I use the 8bit one like this:
The Ollama one supports image inputs too: Output here: https://gist.github.com/simonw/89005e8aa2daef82c53c2c2c62207...Simon, can you recommend some small models that would be usable for coding on a standard M4 Mac Mini (only 16G ram) ?
That's pretty tough - the problem is that you need to have RAM left over to run actual applications!
Qwen 3 8B on MLX runs in just 5GB of RAM and can write basic code but I don't know if it would be good enough for anything interesting: https://simonwillison.net/2025/May/2/qwen3-8b/
Honestly though with that little memory I'd stick to running against hosted LLMs - Claude 3.7 Sonnet, Gemini 2.5 Pro, o4-mini are all cheap enough that it's hard to spend much money with them for most coding workflows.
There are plenty of smaller (quantized) models that fit well on your machine! On a M4 with 24GB it’s already possible to comfortably run 8B quantized models.
Im benchmarking runtime and memory usage for a few of them: https://aukejw.github.io/mlx_transformers_benchmark/
I imagine that this in quantized form would fit pretty well and be decent. (Qwen R1 32b[1] or Qwen 3 32b[2])
Specifically the `Q6_K` quant looks solid at ~27gb. That leaves enough headroom on your 64gb Macbook that you can actually load a decent amount of context. (It takes extra VRAM for every token of context you need)
Rough math, based on this[0] calculator is that it's around ~10gb per 32k tokens of context. And that doesn't seem to change based on using a different quant size -- you just have to have enough headroom.
So with 64gb:
- ~25gb for Q6 quant
- 10-20gb for context of 32-64k
That leaves you around 20gb for application memory and _probably_ enough context to actually be useful for larger coding tasks! (It just might be slow, but you can use a smaller quant to get more speed.)
I hope that helps!
0: https://huggingface.co/spaces/NyxKrage/LLM-Model-VRAM-Calcul...
1: https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-32...
2: https://huggingface.co/Qwen/Qwen3-32B-GGUF
I really like Mistral Small 3.1 (I have a 64GB M2 as well). Qwen 3 is worth trying in different sizes too.
I don't know if they'll be good enough for general coding tasks though - I've been spoiled by API access to Claude 3.7 Sonnet and o4-mini and Gemini 2.5 Pro.
16GB on a mac with unified memory is too small for good coding models. Anything on that machine is severely compromised. Maybe in ~1 year we will see better models that fit in ~8gb vram, but not yet.
Right now, for a coding LLM on a Mac, the standard is Qwen 3 32b, which runs great on any M1 mac with 32gb memory or better. Qwen 3 235b is better, but fewer people have 128gb memory.
Anything smaller than 32b, you start seeing a big drop off in quality. Qwen 3 14b Q4_K_M is probably your best option at 16gb memory, but it's significantly worse in quality than 32b.
With around 4.6 GiB model size the new Qwen3-8B quantized to 4-bit should fit comfortably in 16 GiB of memory: https://huggingface.co/mlx-community/Qwen3-8B-4bit
> I think this is a game changer, because data privacy is a legitimate concern for many enterprise users.
Indeed. At work, we are experimenting with this. Using a cloud platform is a non-starter for data confidentiality reasons. On-premise is the way to go. Also, they’re not American, which helps.
> Btw, you can also run Mistral locally within the Docker model runner on a Mac.
True, but you can do that only with their open-weight models, right? They are very useful and work well, but their commercial models are bigger and hopefully better (I use some of their free models every day, but none of their commercial ones).
I also kind of don't understand how it seems everyone is using AI for coding. I haven't had a client yet which would have approved any external AI usage. So I basically use them as search engines on steroids, but code can't go directly in or out.
Personally I am trying to see if we can leverage AI to help write design documents instead of code, based on a fairly large library of human (poorly) written design documents and bug reports.
You might be able to get your clients to sign something to allow usage, but if you don't, as you say, it doesn't seem wise to vibe code for them. For two reasons:
1. A typical contract transfers the rights to the work. The ownership of AI generated code is legally a wee bit disputed. If you modify and refactor generated code heavily it's probably fine, but if you just accept AI generated code en masse, making your client think that you wrote it and it is therefore their copyright, that seems dangerous.
2. A typical contract or NDA also contains non disclosure, i.e. you can't share confidential information, e.g. code (including code you _just_ wrote, due to #1) with external parties or the general public willy nilly. Whether any terms of service assurances from OpenAI or Anthropic that your model inputs and outputs will probably not be used for training are legally sufficient, I have doubts.
IANAL, and _perhaps_ I'm wrong about one or both of these, in one or more countries, but by and large I'd say the risk is not worth the benefit.
I mostly use third party LLMs like I would StackOverflow: Don't post company code there verbatim, make an isolated example. And also don't paste from SO verbatim. I tried other ways of using LLMs for programming a few times in personal projects and can't say I worry about lower productivity with these limitations. YMMV.
(All this also generally goes for employees with typical employment contracts: It's probably a contract violation.)
> Nobody is seriously disputing the ownership of AI generated code
From what I've been following it seems very likely that, at least in the US, AI-generated anything can't actually be copyrighted and thus can't have ownership at all! The legal implications of this are yet to percolate through the system though.
this is "Kool-aid" from the supply side of LLMs for coding IMO. Plenty of people are plenty upset about the capture of code at Github corral, fed into BigCorp$ training systems.
parent statement reminds me of smug French in a castle north of London circa 1200, with furious locals standing outside the gates, dressed in rags with farm tools as weapons. One well-equipped tower guard says to another "no one is seriously disputing the administration of these lands"
Not sure I can agree with the "I'm billing by the hour" part.
I mean sure, but I think of my little agency providing value, for a price. Clients have budgets, they have limited benefits from any software they build, and in order to be competitive against other agencies or their internal teams, overall, I feel we need to provide a good bang for buck.
But since it's not all that much about typing in code, and since even that activity isn't all that sped up by LLMs, not if quality and stability matters, I would still agree that it's completely fine.
“We” took care to not copy it verbatim (it’s the concrete code form that is copyrighted, not the algorithm), and depending on jurisdiction there is the concept of https://en.wikipedia.org/wiki/Threshold_of_originality in copyright law, which short code snippets on Stack Overflow typically don’t meet.
It's roughly the same, legally, and I was well aware of that.
Legally speaking, you also want to be careful about your dependencies and their licenses, a company that's afraid to get sued usually goes to quite some lengths to ensure they play this stuff safe. A lot of smaller companies and startups don't know or don't care.
From a professional ethics perspective, personally, I don't want to put my clients in that position unless they consciously decide they want that. They hire professionals not just to get work done they fully understand, but to a large part to have someone who tells them what they don't know.
SO seems different because the author of the post is republishing it. If they are republishing copyrighted material without notice, it seems like the SO author is the one in violation of copyright.
In the LLM case, I think it’s more of an open question whether the LLM output is republishing the copyrighted content without notice, or simply providing access to copyrighted content. I think the former would put the LLM provider in hot water, while the latter would put the user in hot water.
Sure there's evidence: Your statements about this when challenged. And perhaps to a degree the commit log, at least that can arouse suspicion.
Sure, you can say "I'd just lie about it". But I don't know how many people would just casually lie in court. I sure wouldn't. Ethics is one thing, it takes a lot of guts, considering the possible repercussions.
I have good results running Ollama locally with olen models like Gemma 3, Qwen 3, etc. The major drawback is slower inference speed. Commercial APIs like Google Gemini are so much faster.
Still, I find local models very much worth using after taking the time to set them up with Emacs, open-codex, etc.
You can set up your IDE to use local LLMs through e.g. Ollama if your computer is powerful enough to run a decent model.
How is it different from the cloud? Plenty startups store their code on github, run prod on aws, and keep all communications on gmail anyway. What's so different about LLMs?
It’s not different. If you have a confidentiality requirements like that, you also don’t store your code off-premises. At least not without enforceable contracts about confidentiality with the service provider, approved by the client.
>How is it different from the cloud? Plenty startups store their code on github, run prod on aws, and keep all communications on gmail anyway. What's so different about LLMs?
Those plenty startups will also use Google, OpenAi or the built in Microsoft AI.
This is clearly for companies that need to keep the sensitive data under their control. I think they also get support with adding more training to the model to be personalized for your needs.
I think that using something like Claude on Amazon Bedrock makes more sense than directly using Anthropic. Maybe I'm naive but I trust AWS more than Anthropic, OpenAI, or Google to not misuse data.
Most my clients have the same requirement. Given the code bases I see my competition generating, I suspect other vendors are simply violating this rule.
I would take any such claim with a heavy rock of salt because the usefulness of AI is going to vary drastically with the sort of work you're tasked with producing.
Are your clients not on AWS/Azure/GCP? They all offer private LLMs out of the box now.
Have you tried using private inference that uses GPU confidential computing from Nvidia?
Game changer feels a bit strong. This is a new entry in a field that's already pretty crowded with open source tooling that's already available to anyone with the time and desire to wire it all up. It's likely that they execute this better than the community-run projects have so far and make it more approachable and Enterprise friendly, but just for reference I have most of the features that they've listed here already set up on my desktop at home with Ollama, Open WebUI, and a collection of small hand-rolled apps that plug into them. I can't run very big models on mine, obviously, but if I were an Enterprise I would.
The key thing they'd need to nail to make this better than what's already out there is the integrations. If they can make it seamless to integrate with all the key third-party enterprise systems then they'll have something strong here, otherwise it's not obvious how much they're adding over Open WebUI, LibreChat, and the other self-hosted AI agent tooling that's already available.
Actually you shouldn't be running LLMs in Docker on Mac because it doesn't have GPU support. So the larger models will be extremely slow if they'll even produce a single token.
I have an M4 Mac Mini with 24GB of RAM. I loaded Studio.LM on it 2 days ago and had Mistral NeMo running in ten minutes. It's a great model, I need to figure out how to add my own writing to it, I want it to generate some starter letters for me. Impressive model.
> Btw, you can also run Mistral locally within the Docker model runner on a Mac.
Efficiently? I thought macOS does not have API so that Docker could use GPU.
I haven't/wouldn't use it because I have a decent K8S ollama/open-webui setup, but docker announced this a month ago: https://www.docker.com/blog/introducing-docker-model-runner
Hmm, I guess that is not actually running inside container/ there is no isolation. Some kind of new way that mixes llama.cpp , OCI format and docker CLI.
I think many in this thread are underestimating the desire of VPs and CTOs to just offload the risk somewhere else. Quite a lot of companies handling sensitive data are already using various services in the cloud and it hasn't been a problem before - even in Europe with its GDPR laws. Just sign an NDA or whatever with OpenAI/Google/etc. and if any data gets leaked they are on the hook.
Good luck ever winning that one. How are you going to prove out a data leak with an AI model without deploying excessive amounts of legal spend?
You might be talking about small tech companies that have no other options.
What's the point when we can run much powerful models now? Qwen3 , Deepseek
It would be short-termist for Americans or euros to use chinese-made models. Increasing their popularity has an indirect but significant cost in the long term. china "winning AI" should be an unacceptable outcome for America or europe by any means necessary.
Why not use confidential computing based offerings like Azure's private inference for privacy concerns?