I chased down what the "4x faster at AI tasks" was measuring:
> Testing conducted by Apple in January 2026 using preproduction 13-inch and 15-inch MacBook Air systems with Apple M5, 10-core CPU, 10-core GPU, 32GB of unified memory, and 4TB SSD, and production 13-inch and 15-inch MacBook Air systems with Apple M4, 10-core CPU, 10-core GPU, 32GB of unified memory, and 2TB SSD. Time to first token measured with an 8K-token prompt using a 14-billion parameter model with 4-bit quantization, and LM Studio 0.4.1 (Build 1). Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Air.
A bit strange to use time to first token instead of throughput.
Latency to the first token is not like a web page where first paint already has useful things to show. The first token is "The ", and you'll be very happy it's there in 50ms instead of 200ms... but then what you really want to know is how quickly you'll get the rest of the sentence (throughput)
As far as benchmarketing goes they clearly went with prefill because it's much easier for apple to improve prefill numbers (flops-dominated) than decode (bandwidth-dominated, at least for local inference); M5 unified memory bandwidth is only about 10% better than the M4.
No you don't. Not as a sticky mushy human with emotions watching tokens drip in. There's a lot of feeling and emotion not backed by hard facts and data going around, and most people would rather see something happening even if it takes longer overall. Hence spinner.gif, that doesn't actually remotely do a damned thing, but it gives users reassurance that they're waiting for something good. So human psychology makes time to first token an important metric to look at, although it's not the only one.
"Scaling up performance from M5 and offering the same breakthrough GPU architecture with a Neural Accelerator in each core, M5 Pro and M5 Max deliver up to 4x faster LLM prompt processing than M4 Pro and M4 Max, and up to 8x AI image generation than M1 Pro and M1 Max."
Are they doubling down on local LLMs then?
I still think Apple has a huge opportunity in privacy first LLMs but so far I'm not seeing much execution. Wondering if that will change with the overhaul of Siri this spring.
I think its just marketing, and the marketing is working. Look how many people bought Minis and ended up just paying for API calls anyway. (Saw it IRL 2x, see it on reddit openclaw daily)
I don't mind it, I open Apple stock. But I'm def not buying into their rebranding of integrated GPU under the guise of Unified Memory.
> Look how many people bought Minis and ended up just paying for API calls anyway. (Saw it IRL 2x, see it on reddit openclaw daily)
Aren't the OpenClaw enjoyers buying Mac Minis because it's the cheapest thing which runs macOS, the only platform which can programmatically interface with iMessage and other Apple ecosystem stuff? It has nothing to do with the hardware really.
Still, buying a brand new Mac Mini for that purpose seems kind of pointless when a used M1 model would achieve the same thing.
It’s exactly that. They are buying the base model just for that. You are not going to do much local AI with those 16GB of ram anyway, it could be useful for small things but the main purpose of the Mini is being able to interact with the apple apps and services.
16GB should be enough for TTS/Voice models running locally no ? I was thinking about having a home assistant setup like that where the voice is local and the brain is API based
No one is buying a base model Mac for local LLM. Everyone is forgetting the PC prices have drastically increased due to RAM and SSD. Meanwhile, Macs had no such price change… at least for the models that didn’t just drop today. Mac’s are just a good deal at the moment.
Yeah because Mac upgrade prices were already sky high, long before the component shortage. 32GB of DDR5-6000 for a PC rocketed from $100 to $500, while the cost of adding 16GB to a Mac was and still is $400.
Worse than that, they hold their value, so buying a used M1 mini is still a few hundred bucks, and saving $200-300 by purchasing a 5 generation older mini seems like a bad deal in comparison.
> Aren't the OpenClaw enjoyers buying Mac Minis because it's the cheapest thing which runs macOS
That's likely only part of the reason. Mac Mini is now "cheap" because everyone exploded in price. RAM and SSD etc have all gone up massively. Not the mention Mac mini is easy out of the box experience.
It's not cheap, though. Two weeks ago I bought a computer with a similar form factor (GMKtec G10). Worse CPU and GPU but same 16GB memory and a larger SSD for 40% the price of a base mac mini ($239 vs $599). It came with Windows preinstalled, but I immediately wiped that to install linux. Even a used (M-series) mac mini is substantially more expensive. It will cost me about an extra penny per day in electricity costs over a mac mini, but I won't be alive long enough for the mac mini to catch up on that metric.
I considered the mac mini at the time, but the mac mini only makes sense if you need the local processing power or the apple ecosystem integration. It's certainly not cheaper if you just need a small box to make API calls and do minimal local processing.
If you just need "a small box to make API calls and do minimal local processing" you an also just buy a RPI for a fraction of the price of the GMKtec G10.
All 3 serve a different purpose; just because you can buy a slower machine for less doesn't mean the price:performance of the M1 Mac Mini changes.
Why not? The integrated GPUs are quite powerful, and having access to 32+ GB of GPU memory is amazing. There's a reason people buy Macs for local LLM work. Nothing else on the market really beats it right now.
My M4 MacBook Pro for work just came a few weeks ago with 128 GB of RAM. Some simple voice customization started using 90GB. The unified memory value is there.
Jeff Geerling had a video of using 4 Mac Studios each with 512GB RAM connected by Thunderbolt. Each machine is around $10K so this isn't cheap but the performance is impressive.
It’s what a small business might have paid for an onprem web server a couple of decades ago before clouds caught on. I figure if a legal or medical practice saw value in LLMs it wouldn’t be a big deal to shove 50k into a closet
I've tried to use a local LLM on an M4 Pro machine and it's quite painful. Not surprised that people into LLMs would pay for tokens instead of trying to force their poor MacBooks to do it.
Local LLM inference is all about memory bandwidth, and an M4 pro only has about the same as a Strix Halo or DGX Spark. That's why the older ultras are popular with the local LLM crowd.
And while it is stupid slow, you can run models of hard drive or swap space. You wouldn’t do it normally, but it can be done to check an answer in one model versus another.
We had a workshop 6 months ago and while I've always been sceptical of OpenAI,etc's silly AGI/ASI claims, the investments have shown the way to a lot of new technology and has opened up a genie that won't be put back into the bottle.
Now extrapolating in line with how Sun servers around year 2000 cost a fortune and can be emulated by a 5$ VPS today, Apple is seeing that they can maybe grab the local LLM workloads if they act now with their integrated chip development.
But to grab that, they need developers to rely less on CUDA via Python or have other proper hardware support for those environments, and that won't happen without the hardware being there first and the machines being able to be built with enough memory (refreshing to see Apple support 128gb even if it'll probably bleed you dry).
I certainly only use Macs when being project assigned, then there are plenty of developers out there whose job has nothing to do with what Apple offers.
Also while Metal is a very cool API, I rather play with Vulkan, CUDA and DirectX, as do the large majority of game developers.
Honestly though, gamedevs really are among the biggest Windows stalwarts due to SDK's and older 3d software.
Only groups of developers more tied to Windows that I can think of are probably embedded people tied due to weird hardware SDK's and Windows Active Directory dependent enterprise people.
Outside of that almost everyone hip seems to want a Mac.
That's the broad developer community. 90%+ of the engineers at Big Tech and the technorati startups are on MacOS with 5% on Linux and the other 5% on Windows.
You’ll see a lot of MacBooks in Beijing’s zhongguangcun where all the tech companies are, but they also have a lot of students there as well, so who knows. You need to go out to the suburbs where Lenovo has offices to stop seeing them. I know Apple is common in Western Europe having lived there for two years (but that was 20 years ago, I lived in China for 9 years after that).
It wouldn’t surprise me if the deepseek people were primarily using Mac’s. Maybe Alibaba might be using PCs? I’m not sure.
I think it's reasonable to say that the people responding to surveys on Stack Overflow aren't the same people who work on pushing the state of the art in local LLM deployment. (which doesn't prove that that crowd is Apple-centric, of course)
I think that might be partly because on regular PC's you can just go and buy an NVidia card insteaf of fuzzing around with software issues, and for those on laptops they probably hope that something like Zluda will solve it via software shims or MS backed ML api's.
Basically, too many choices to "focus on" makes non a winner except the incumbent.
Except CUDA feels really cozy, because like Microsoft, NVidia understands the Developers, Developers, Developers mantra.
People always overlook that CUDA is a polyglot ecosystem, the IDE and graphical debugging experience where one can even single step on GPU code, the libraries ecosystem.
And as of last year, NVidia has started to take Python seriously and now with cuTile based JIT, it is possible to write CUDA kernels in pure Python, not having Python generate C++ code that other tools than ingest.
Neural Accelerators (aka NAX) accelerates matmults with tile sizes >= 32. From a very high level perspective, LLM inference has two phases: (chunked) prefill and decode. The former is matmults (GEMM) and the latter is matrix vector mults (GEMV). Neural Accelerators make the former (prefill) faster and have no impact on the latter.
There already are a bunch of task-specific models running on their devices, it makes sense to maintain and build capacity in that area.
I assume they have a moderate bet on on-device SLMs in addition to other ML models, but not much planned for LLMs, which at that scale, might be good as generalists but very poor at guaranteeing success for each specific minute tasks you want done.
In short: 8gb to store tens of very small and fast purpose-specific models is much better than a single 8gb LLM trying to do everything.
> I still think Apple has a huge opportunity in privacy first LLMs
This correlation of Apple and privacy needs to rest. They have consistently proven to be otherwise - despite heavily marketing themselves as "privacy-first"
No other company makes you tell them every application you install on your device. No other company makes you tell them every location you read from your GPS sensor.
Not for everything. Apple has initially focused on edge AI that runs locally per device. It didn’t work out well the first try, but I would still bet on them trying again once compute catches up. Besides, they still have a better track record than the other tech giants.
The hardware capabilities that make local LLMs fast are useful for a lot of different AI workloads. Local LLMs are a hot topic right now so that’s what the marketing team is using as an example to make it relatable.
But memory bandwidth (bottleneck for LLM inference) is only marginally improved, 614 GB/s vs 546 GB/s for M4/M5 Max - where is this 4x improvement coming from?
What % of users actually care that much about local LLMs? It appears to still be an inferior (though maybe decent) service compared to ChatGPT etc., and requires very top-end hardware. Is privacy _that_ important to people when their Google search history has been a gateway to the soul for years? I wonder if these machines would cost significantly less (or put the cost to other things, e.g. more CPU cores) without this emphasis on LLMs.
Apple's AI strategy really kind of threads the needle cleverly.
"AI" (LLMs) may or may not have a bubble-pop moment, but until it does Apple get to ride it on these press releases and claims. But if the big-pop occurs, then Apple winds up with really fantastic hardware that just happens to be good at AI workloads (as well as general computing).
For example, image classification (e.g. face recognition/photo tagging), ASR+vocoders, image enhancement, OCR, et al, were popular before the current boom, and will likely remain popular after. Even if LLM usage dries up/falls out of vogue, this hardware still offers a significant user benefit.
They could run fine on the CPU too. But these are mobile devices, therefore battery usage is another significant metric. Dedicated hardware is more energy efficient than general hardware, and GPU in particular is a power-hog.
Exactly. It's the same thing as video or audio encoding and decoding. Sure the CPU could do it, potentially use the GPU, but having actual hardware encoders and decoders for the most common codecs will save a lot of energy.
Honestly, I think that's the move for apple. They do not seem to have any interest in creating a frontier lab/model -- why would they give the capex and how far behind they are.
But open source models (Kimi, Deepseek, Qwen) are getting better and better, and apple makes excellent hardware for local LLMs. How appealing would it be to have your own LLM that knows all your secrets and doesnt serve you ads/slop, versus OpenAI and SCam Altman having all your secrets? I would seriously consider it even if the performance was not quite there. And no need for subscription + cli tool.
I think apple is in the best position to have native AI, versus the competition which end up being edge nodes for the big 4 frontier labs.
Apple absolutely has a massive opportunity here because they used a shared memory architecture.
So as most people in or adjacent to the AI space know, NVidia gatekeeps their best GPUs with the most memory by making them eye-wateringly expensive. It's a form of market segmentation. So consumer GPUs top out at 16GB (5090 currently) while the best AI GPUs (H200?) is 141GB (I just had to search)? I think the previou sgen was 80GB.
But these GPUs are north of $30k.
Now the Mac Studio tops out currently at 512GB os SHARED memory. That means you can potentially run a much larger model locally without distributing it across machines. Currently that retails at $9500 but that's relatively cheap, in comparison.
But, as it stands now, the best Apple chips have significantly lower memory bandwidth than NVidia GPUs and that really impacts tokens/second.
So I've been waiting to see if Apple will realize this and address it in the next generation of Mac Studios (and, to a lesser extend, Macbook Pros). The H200 seems to be 4.8TB/s. IIRC the 5090 is ~1.8TB/s. The best Apple is (IIRC) 819GB/s on the M3 Ultra.
Apple could really make a dent in NVidia's monopoly here if they address some of these technical limitations.
So I just checked the memory bandwidth of these new chips and it seems like the M5 is 153GB/s, M5 Pro is ~300 and M5 Max is ~600. I was hoping for higher. This isn't a big jump from the M4 generation. I suspect the new Studios will probably barely break 1TB/s. I had been hoping for higher.
Hard to get 6000+ bit memory bus HBM bandwidth out of a 512 or 1024 bit memory bus tied to DDR... I think it's also just tough to physically tie in 512 gigs close enough to the GPU to run at those speeds. But yeah, I wish there was a very competitive local option, too, short of spending $50k+.
It is simply marketing nonsense - what they really mean (I think) is they support matrix multiplication (matmul) at the hardware level which given AI is mostly matrix multiplications you'll get much faster inference (and some increase in training too) on this new hardware. I'm looking forward to seeing how fast a local 96gb+ LLM is on the M5 Max with 128gb of RAM.
It’s not necessarily doubling down on local. The reality is your LLM should be inferencing every tick … the same way your brain thinks every. Fucking. Nano. Second.
So yes, the LLM should be inferencing on your prompt, but it should also be inferencing on 25,000 other things … in parallel.
Those are the compute needs.
We just need compute everywhere as fast as possible.
A useful llm that needs 64gb of ram and mid double digit cores is not useful for 99% of their customers. The LLMs they have on iphone 17's certainly cannot do anything useful other than summerization and stuff. It's a hardware constraint that they have.
The topic is MacBook, so my criticism is a little off. However, I really dont believe in this "local LLM" promise from Apple. My phone already gets noticeably warm if I answer 5 WhatsApp messages. And looses 5% of battery during the process. I highly doubt Apple will have a useable local LLM that doesn't drain my battery in minutes, before 2030.
Something is not right if WhatsApp is seriously draining your phone like that. Admittedly I’m not a big WhatsApp user my iPhone hasn’t had any trouble like that with it.
I've been so disappointed in Apple's lack of execution on this. There is so much potential for fantastic local models to run and intelligently connect to cloud models.
I just don't get why they're dropping the ball so much on this.
I typed “RAM” to search for it and boy they hammer home how lucky I am to be getting 1TB SSD standard, but no mention of RAM anywhere on this page. Anyway, the MacBook Pro starts with 16GB of RAM. It’s $400 to go from 16GB to 32GB.
Interestingly, 36-128GB models are showing as “currently unavailable” on the store page, and you can’t even place an order for them right now? But for anyone curious, it’s quoting $5099 for the 128GB RAM 14” MacBook Pro model.
No change from the previous models then, 16GB->32GB was already $400. They're cutting into their previously enormous margins to keep the prices stable, rather than hiking the prices to maintain their margins.
They bought the fab time for that RAM 2-3 years ago. Apple is renowned for their foresight and preparation. We'll eventually see price increases from Apple's RAM upgrade, but we're not there yet.
This is not exactly correct. If you have an M5 Pro chip instead of m5 Chip - I just built a 16inch, M5 Pro chip, it is $400 to go from 24 -> 48gb. An additional $200 ($600 over base) to go to 64gb. So the memory prices change based on chip. M5 Max Chip starts with 48gb of memory.
I know RAM is scarce and everything, but doubling down on LLM local acceleration with all of that dedicated silicon while at the same time sticking with Apple's traditional lack of RAM availability makes for a very weird product proposition to me.
> M5 Pro supports up to 64GB of unified memory with up to 307GB/s of memory bandwidth, while M5 Max supports up to 128GB of unified memory with up to 614GB/s of memory bandwidth
Ah yeah you’re right, thanks. I tried to at least make my post useful and pull up prices for the different tiers. Overall, those prices are surprisingly competitive now compared to the rest of the laptop market!
Fair chance that Apple has price/purchase agreements already in place. Consumers are left to fight over the excess capacity after megabuyers get their orders filled.
> Interesting that this hasn't budged since the memory shortages appeared.
Apple has had enough war chests with the ability of buying the entirety of TSMC's new capacity years in advance in the past.
If I were to guess, Apple locked in their entire BOM and production capacity two years ago. That's something even the large players cannot replicate because they run cash-lean and have too many different SKUs, and the small players (Framework, System76, even Steam) are entirely left to the forces of the markets.
More to do with the faster storage allowing you to swap without noticing it as much. There was this whole trend when m1 first came out of people saying it didn't matter if you got the lowest spec because the ssd was so fast it made up for the lack of ram... totally ignoring that swapping like that was destroying their drives really fast.
My wife’s 8GB MacBook Air crashed yesterday with Firefox and Find My open and nothing else because of running out of RAM, so, sort of, but they’re not magic. (Find My was using 3GB of memory!)
It’s a losing battle me trying to tell my wife to close her Firefox tabs, haha, but yes, Firefox does use a lot of ram when you have 500 tabs. Maybe I’ll get her a 64GB MacBook Pro for the premium web browsing experience she so desires!
You mean bumped $100. M4 MacBook Pro and M5 MacBook Pro started at $1599 with 512GB SSD.
Now it starts at $1699, a $100 bump but comes with a 1TB SSD. Previously it would have cost $1799 for the 1T SSD, so it's a $100 bump on base price but you are also getting 1TB SSD for $100 less than before.
Except that it's literally not true and people are repeating it for some stupid reason, I assume you just never actually looked it up - laptops are specifically excluded from that regulation, and in fact Apple does bundle a power adapter with their laptops, just not on the cheapest models.
To be fair, ever since the advent of high power USB-C PD that really, really is not needed any more, way too many power bricks are effectively e-waste.
People already have USB-C power bricks and docks everywhere and unlike pre-USB-C generations, you can use them not just across different generations of hardware, but across vendors as well.
> The new 14- and 16-inch MacBook Pro with M5 Pro and M5 Max mark a major leap for pro users. There’s never been a better time for customers to upgrade from a previous generation of MacBook Pro with Apple silicon or an Intel-based Mac.
I read as "Whoops we made the M1 Macbook Pro too good, please upgrade!"
I think I will get another 2-5 years out my mine.
Apple: If you document the hardware enough for the Asahi team to deliver a polished Linux experiene, I'll buy one this year!
My 32gb m1 max was probably the best purchase I've made. Still plenty of headroom in performance left in this beast. Wonder what reason they'll use to end software support in the future. Bet it'll be some security hardware they make up for the sake of forcing upgrades.
my tinfoil hat theory is that they make small features depend on new hardware.
for example, let's say the new os depends on m5's exclusive thumbnail generator accelerator, and let's say it improves speed by a 20%.
now, your M1 notebook than on previous OSes uses standard gpu acceleration for thumbnails will not have this specialized hardware acceleration, it will have software fallback that will be 90% slower.
you won't notice it a first thought because it's stuff, fast, but it eats a bit of the processor.
multiply this by 1000 features and you have a slow machine.
I don't know how else to explain how an ipad pro cannot even scroll a menu without stuttering, it's insane how fast these things were on release
yes pretty much this. make useless features use up resources and make basic scrolling slow.
the Liquid Glass for example probably is not so great when it comes to resources. Probably works better with latest metal and hardware blocks on the GPU in M5 as opposed to using GPU cores and unified memory on 8gb M1 making latest macOS work not so great. I have the M1 8gb air and it is really slow on Tahoe. It was snappy just a couple of years ago on a fresh install.
I've been on a Macbook M1 Pro since 2022 (bought refurbished on Amazon for cheap) and it's still such a powerhouse. It doesn't struggle at all with anything that I throw at it. Kind of amazing.
Nothing has broken and I consistently get 4-6 hours of heavy work time while on battery. An amazing machine for the price I paid.
Same, in fact the only reason right now that I would upgrade my m1 pro is if they threaten to change the design by getting rid of the hdmi or sd card slot, or doing something stupid like when they added the touch bar. I was locked into my old intel pro for so long because of all the bad hardware choices they were making.
I read it the same way. I should've gotten way more RAM back when I got my M1 and RAM was still cheap although this was of course before the LLM boom so there was no way to really know.
I maxed my M1 out when I bought it because I was frustrated with the 16GB max in the previous machines. I use my machine for all sorts of things and some days you just don't feel like exiting apps to make space for new ones.
I still don't have a strong urge to upgrade. I could probably get by on 32GB (like my work-issued machine is) but 64GB is the right amount of headroom for me.
There was a magical window at Google where you could be issued an iMac Pro 5k. (To this day, the standard issue monitor is still 1440p.)
~9 years later, there are a lot of people still using it as their main machine, waiting until we get kicked off the corp network for lack of software support.
I feel like Apple pulled an Instant Pot with the M1 MacBook Pro. I still haven't had a single situation where I felt like spending more money would improve my experience. The battery is wearing out a bit, but it started out life with so much runtime that losing a few hours doesn't seem to matter.
On M4 Max 128GB we're seeing ~100 tok/s generation on a 30B parameter model in our from scratch inference engine. Very curious what the "4x faster LLM prompt processing" translates to in practice. Smallish, local 30B-70B inference is genuinely usable territory for real dev workflows, not just demos. Will require staying plugged in though.
The memory bandwith on M4 Max is 546 GB/s, M5 Max is 614GB/s, so not a huge jump.
The new tensor cores, sorry, "Neural Accelerator" only really help with prompt preprocessing aka prefill, and not with token generation. Token generation is memory bound.
Hopefully the Ultra version (if it exists) has a bigger jump in memory bandwidth and maximum RAM.
The marketing subterfugue might be about this exactly, technically prompt processing means the prefill phase of inference. So prompt goes in 4x as fast but generates tokens slower.
This seems even likely as the memory bandwidth hasn't increased enough for those kinds of speedups, and I guess prefill is more likely to be compute-bound (vs mem bw bound).
I strongly agree. People see local "GPT-4 level" responses, and get excited, which I totally get. But how quickly is the fall-off as the context size grows? Because if it cannot hold and reference a single source-code file in its context, the efficiency will absolutely crater.
That's actually the biggest growth area in LLMs, it is no longer about smart, it is about context windows (usable ones, note spec-sheet hypotheticals). Smart enough is mostly solved, combating larger problems is slowly improving with every major release (but there is no ceiling).
That should be covered by the harness rather than the LLM itself, no? Compaction and summarization should be able to allow the LLM to still run smoothly even on large contexts.
4x faster is about token prefill, i.e. the time to first token. It should be on par with DGX Spark there while being slightly faster than M4 for token generation. I.e. when you have long context, you don't need to wait 15 minutes, only 4 minutes.
I run gpt-oss 120b model on ollama (the model is about 65 GB on disk) with 128k context size (the model is super optimized and only uses 4.8 GB of additional RAM for KV cache at this context size) on M4 Max 128 GB RAM Mac Studio and I get 65 tokens/s.
I find time to first token more important then tok/s generally as these models wait an ungodly amount of time before streaming results. It looks like the claims are true based on M5: https://www.macstories.net/stories/ipad-pro-m5-neural-benchm... so this might work great.
>On M4 Max 128GB we're seeing ~100 tok/s generation
You mean on your first token. Whats the performance after 500 and 3000 tokens?
I genuinely don't understand why people post stuff like this. People are not informed enough to know you mean first tokens. They are going to make a mistake and buy one thinking they will get 100tk/s.
Are you working for Apple marketing? Do you have post purchase regret? I cannot imagine deliberately misleading people. Maybe you are hoping more buyers build up your ecosystem?
For chat type interactions prefill is cached, prompt is processed at 400tk/s and generation is 100-107tk/s, it's quite snappy. Sure, for 130,000 tokens, processing documents it drops to, I think 60tk/s, but don't quote me on that. The larger point is that local LLMs are becoming useful, and they are getting smarter too.
I'm not sure if you're just unaware or purposefully dense. It's absolutely possible to get those numbers for certain models in a m4 max and it's averaged over many tokens, I was just getting 127tok/s for 700 token response on a 24b MoE model yesterday. I tend to use Qwen 3 Coder Next the most which is closer to 65 or 70 tok/s, but absolutely usable for dev work.
I think the truth is somewhere in the middle, many people don't realize just how performant (especially with MLX) some of these models have become on Mac hardware, and just how powerful the shared memory architecture they've built is, but also there is a lot of hype and misinformation on performance when compared to dedicated GPU's. It's a tradeoff between available memory and performance, but often it makes sense.
LM Studio (which prioritizes MLX models if you're on Mac and they are available) - I have it setup with tailscale running as a server on my personal laptop. So when I'm working I can connect to it from my work laptop, from wherever I might be, and it's integrated through the Zed editor using its built in agent - it's pretty seamless. Then whenever I want to use my personal laptop I just unload the model and do other things. It's a really nice setup, definitely happy I got the 128gb mbp because I do a lot of video editing and 3d rendering work as a hobby/for fun and it's sorta dual purpose in that way, I can take advantage of the compute power when I'm not actually on the machine by setting it up as a LLM server.
This. I have been a big (and loud) fan of M-series hardware from the beginning, but if Apple is going to keep making their software worse, I will find myself lingering on older generations that run Asahi Linux or going back to a traditional x86_64 laptop instead of buying into new generations.
Same here. I know some people are unhappy with some of the UX tweaks but honestly I don't notice much of it. The whole liquid glass thing is a bit gimmicky. Other than that, I don't see much difference. The rounded corners on windows are a bit silly. But I don't spend a lot of time fiddling with windows. Most of my windows are maximized (not full screen). I'm sure there are other issues people dislike that I just haven't noticed.
I use my laptop for development. I don't actually use most of the built in applications. My browser is Firefox, I use codex, vs code, intellij, iterm2, etc. Most of that works just fine just as it did on previous versions of the OS. I actually on purpose keep my tool chains portable as I like to have the option to switch back to Linux when I want to. I've done that a few times. I come back for the hardware, not the OS.
In my experience, if you don't like Apple's OS changes that is unfortunate but they don't seem to generally respond to a lot of the criticism. Your choices are to get further and further out of date, switch to something else, or just swallow your pride. Been there done that. Windows is a "Hell No" for me at this point. I'll take the UX, with all the pastel colors that came and went and all the other crap that got unleashed on macs over the last ten years. Definitely a case of the grass not being greener on Windows. Even with the tele tubby default desktop in XP back in the day.
I can deal with Linux (and use that on and off on one of my laptops). However, that just doesn't run that well on mac hardware. And any other hardware seems like a big downgrade to me. Both Windows and Linux are arguably a lot worse in terms of UX (or lack thereof). Linux you can tweak. And you kind of have to. But it just never adds up to consistent and delightful. Windows, well, at this point liking that is probably a form of Stockholm Syndrome. If that doesn't bother you, good for you.
So, Mac OS it is for me as everything else is worse. I've in the past deferred updates to new versions of Mac OS as well. Generally you can do that for a while but eventually it becomes annoying when things like homebrew and other development toys start assuming you run something more recent. And of course for security reasons you might just not drag your feet too long. Just my personal, pragmatic take.
Closing Tabs in Safari till takes more than a second though. And if you hold Cmd-W to close all of them it just completely locks up and crashes. Still not fixed since the release of Safari 26.
I will say that 26.4 beta 2 was the first time I've regretting using betas since Sonoma beta 2. The Sonoma beta ruined the firmware on my machine and Apple had to replace the logic board; the latest Tahoe beta broke all networking on my machine and I had to erase the installation to fix everything. I've since dropped off the beta train for the time being.
I already left the beta train on my iPhone because I had too many issues getting my grocery apps to allow me to place orders without going to my laptop and doing it in a web browser.
I'm on an M4 Pro MacBook-- basically the fastest computer you could buy from Apple before today-- and opening/closing the tab sidebar in Safari on Tahoe takes multiple seconds, even if I have only 4-6 tabs open, and seems to drop to 5 FPS. It's comically bad.
It's so bad I switched back to Chrome. I had thought Chrome had a major battery life penalty compared to Safari on Macs, but I checked more up-to-date info and apparently that's outdated.
I moved away from mac because of the OS and couldn't be happier. The hardware may be great but non-Apple hardware is fine too, and Linux is significantly better experience than MacOS these days.
The next macOS will be touch screen centric with elements getting bigger when you're close to touching them, rumors say. That being said, I run Tahoe and it works perfectly fine to me, I am not sure what issues people have with it. Sure, some corner radii aren't exactly the same but I honestly couldn't give less of a shit as long as it runs the programs I need.
Safari routinely using 20+ Gb of memory with a handful of tabs open. Safari tabs refusing to close. Unresponsive System Settings window. Random application freezes and crashes, Apple Music not playing music. This is on a 32Gb M1 Max. My M1 Air on Sequoia doesn't experience any of these issues, even if it has half the unified memory.
Not necessarily, because I never used Apple apps, it's not like I'm avoiding them now because they're ostensibly buggy (as others don't seem to have the same issues in this thread).
I don't think the "new" Performance cores are just "renamed" "E" / "Efficiency" cores; Apple has retroactively renamed the baseline M5 nomenclature to say it has "10-core CPU with 4 super cores and 6 efficiency cores"; so they're clearly keeping the "efficiency cores" nomenclature around.
I think this is a new design, with Apple having three tiers of cores now, similar to what Qualcomm has been doing for a while.
I think how it breaks down is:
- "Super" are the old "P" cores, and the top tier cores now
- "Performance" cores are a new tier and seen for the first time here, slotting between "old" P and E in performance
- "Efficiency" / "E" are still going to be around; but maybe not in desktop/Pro/Max anymore.
Interesting. This is clearly a big CPU change if so. I wonder why no E cores. I’m sure E cores would be more efficient at OS tasks than the new performance cores.
For example, 6 super, 8 performance, and 4 efficiency.
I was looking into this. The M5 performance cores can be scaled down to match efficiency cores in performance and power usage.
I believe they lower the clock speed, limit how much work is done in parallel on each core, and limit how aggressive the speculative execution is so less work is wasted.
> The industry-leading super core was first introduced as performance cores in M5, which also adopts the super core name for all M5-based products
But new "performance" is claimed to be new design (= not just overclocked efficiency core from M5?):
> M5 Pro and M5 Max also introduce an all-new performance core that is optimized to deliver greater power-efficient, multithreaded performance for pro workloads.
If there’s anything this past three years has taught me, it’s that modern cpus can performantly do every task except for streaming text over the internet.
I've found current-generation Macs so capable that I've switched to using a Macbook Air. Would strongly recommend - it's still a powerful machine and it's significantly lighter and cheaper.
I have a powerful older Mac that doesn’t really “choke” on anything, but I could always use more speed.
The high memory Macs have been great for being able to run LLMs, but the prompt processing has always been on the slow side. The new AI acceleration in these should help with that.
There are also workloads like compiling code where I’ll take all the extra speed I can get. Every little bit of reduced cycle time helps me finish earlier in the day.
And then there’s gaming. I don’t game much, but the M1 and M2 era Apple Silicon feels sluggish relative to what I have on the nVidia side.
AI video generation can fairly easily choke anything that's not NVIDIA's flagship model. Even the latest local image gen models are so large that they can be frustratingly slow with non-optimal hardware even if they fit in the VRAM. IIRC when I had an M2, it was about 4x slower at running the venerable Stable Diffusion (and SDXL) than my meager RTX 3060.
Sounds pretty beefy. What kind of local LLM is that thing capable of running? Does it open up real alternatives to cloud providers like OpenAI and Claude, or are the local models this hardware is capable of running still pretty far behind?
You might have confused Hacker News with your e-mail inbox again. This is an Apple press release, directed to everybody in the world who might be interested in a new computer or their first computer.
What’s with the attitude? My machine is aging like a fine wine, I’m acknowledging how resilient their custom silicon is despite the world demanding more and more compute.
It was a joke, should have put a smiley face. But every thread on a new Apple product here on HN have the same "why should I upgrade" comment, forgetting that there are people who might have very old devices they want to upgrade, or they might want to switch from Windows/Android to Apple.
Even if a new device is a small upgrade from last year's model, it can be a giant upgrade for other people.
Are you one of the folks thinking of upgrading? If so, from what generation? What makes you excited? Isn't this a more interesting way to have the conversation?
Got it. I guess it feels unfair to gaslight people who are celebrating not needing upgrades, anecdotally sharing their experiences - because some people just need a new computer for xyz reason in time.
Interesting that they're showing VFX/CG software (Autodesk MAYA and Foundry Nuke) so prominently - obviously people using "Pro" machines are the target audience for this, but both of those apps (any many others in the industry) use Qt for the interface, rather than being totally platform-native.
Contrary to HN popular belief, there are neither incentives nor benefits to building native ui apps, for neither consumer nor professional apps. The exception is apps that only make sense on a single platform, such as window management and other deep integration. On iOS/macos you have a segment of indie/smaller apps that capture a niche market of powerusers for things like productivity apps. But the point is it makes no sense for anything from Slack, VSCode, Maya, DaVinci Resolve, and so on, to build native UIs. Even if they wanted to build and maintained 3 versions, advanced features aren’t always available in these frameworks. In the case of Windows, even MS has given up on their own tech, and have opted to launch webview based apps. Apple is slightly more principled.
Strong disagree. I think Microsoft’s decision to wrap web apps for the desktop is one of the stupidest they have ever made. It provides poor user experience, uses more battery power and needs more memory and CPU to be performant and creates inconsistencies and wierd errors compared to native apps.
$200 price bump across the board. The cheapest 16" is now $2699 and 14" Pro $2199. I think it's a fair price considering M2Pro 14" was $1999 (though it was discounted) only had 512GB and 16GB RAM.
It's not $200 across the board. M4 MacBook Pro and M5 MacBook Pro started at $1599 with 512GB SSD.
Now it starts at $1699, a $100 bump but comes with a 1TB SSD. Previously it would have cost $1799 for the 1T SSD, so it's a $100 bump on base price but you are also getting 1TB SDD for $100 less than before.
To clarify, I meant, model with Pro chip, not just Macbook Pro name.
For example, up until MacBookPro M2, MacBookPro M2 came with M2 Pro chip.
However, starting with M3, Apple lowered the MacBookPro MSRP to $1599, but its base configuration was downgraded to M3 chip from M3 Pro. To get the M3 Pro, you had to pay $1999. There's substantial performance between the two.
Same with M4. To get the M4 Pro chip, you had to pay $1999.
Now to get M5 Pro chip, it's $2199. Still a good value, but just saying it's a deviation from the trend.
Honest question. Is it possible to install an earlier version of macOS on these machines?
Liquid glass looks so.. unprofessional to my eyes. And I hear it's also unstable.
Considering these max out at 128GB of unified ram my guess is the hope of an M5 Ultra with 1TB of unified ram is unlikely to come true... Super disappointing.
I checked the fine print on the product website: by “up to 4x faster LLM prompt processing,” they’re specifically referring to time to first token. So it’s not about token generation rate (tokens per second).
Yes. This is known. They added neural accelerators, aka Tensor core equivalent, in the GPU. This will make prompt processing competitive vs similar class GPUs.
"The new MacBook Pro gets up to 24 hours of battery life, giving Intel-based upgraders up to 13 additional hours"
I have a Intel-based 2019 Macbook Pro still and I have NEVER in its lifetime gotten even half of what they are claiming here. These days if I run it from battery I might get 90 mins.
That said I had a maxed out Macbook Pro M4 Max on order but just cancelled it right now and will get this new M5 Max one for basically the same price. Once I saw that they didn't up the price of memory (I don't know how it doesn't affect them) I canceled my order.
I had intel MacBook Pro. It is a NIGHT and DAY difference. I wish I didn’t get the 16gb of memory though. It is ok, but running 5-10 cursor ai agents at the same time does start to choke the memory.
Battery is absolutely amazing! And the best part - it stays cold!! No more irritated from heat fingers when using touchpad.
For those who don't already know, you can get a lot of PC gaming performance out of these machines using Sikarugir. You can install all of Steam via winetricks and go from there, or launch DRM-free games directly.
MacBook Pro with M5 Pro now comes standard with 1TB of storage, while MacBook Pro with M5 Max now comes standard with 2TB. And the 14-inch MacBook Pro with M5 now comes standard with 1TB of storage.
Because the M1 was too good, a qualitative leap over previous Macs and really every other laptop and even some desktops back in 2020. Now, Apple Silicon is just iterative.
Me either. I guess it's just fatigue, at least for me. I also don't really get that excited by new LLM releases either. Not to say the tech isn't impressive, but I guess all the hype has me inured.
For me going way back, it was exciting when I had to save a bit (but not too much!) for a new 512 DIMM, and when I opened the box and smelled the chip smell, put it in always worried I was going to fuck it up, and then computer literally felt faster that next boot...that was pretty fun!! Now it's like oh great $5k for a slab of stone that can do pretty much anything, neat. I still think computers are cool, just not particularly exciting.
Because it's the same shit every year for the past 5 years with the M line. 2010 to 2015 was a major improvement, 2015 to 2020 was a major improvement, now they pretty much solved the computer/laptop problem for 99% of people. I'm on a 16gb m1 air, I see absolutely no reason to update.
I was considering it but got cold feet when I've been told that you could damage it when cleaning it. When I open/close my laptop I leave a ton of finger prints. I'm not too good with delicate hardware stuff.
> M5 Pro supports up to 64GB of unified memory with up to 307GB/s of memory bandwidth, while M5 Max supports up to 128GB of unified memory with up to 614GB/s of memory bandwidth.
This is the important statement. 614GB/s is quite decent, however a NVIDIA RTX 5090 already offers 1,792 GB/s (roughly 3x) of memory bandwidth, for comparison.
You're right a $3600 graphics card is worse than a $2600 laptop; but from my perspectives they're very different products. Not least of all because even at $3600 for a RTX 5090 you still have the whole rest of the computer left to purchase.
The RTX 5090 only has 32gb of VRAM. So the tradeoff is NVIDIA is for blazing speed in a tiny memory pool, but Apple Silicon has a larger memory pool at moderate speed.
That's a fun comparison, but can you run those 2 m5 pros in parallel to accomplish 2x the work? Otherwise, you just told me you can buy 2 toyota corollas for the price of 1 F-150 while trying to convince me you can haul your boat behind both corollas at the same time.
All I'm saying is that the comparison doesn't make sense. The 5090 is faster on a small subset of tasks if attached to a computer which ends up being 3x the price of a m5 machine that fit the same model or the same price as a machine that fits models 5x bigger
So you're saying that buying 2 Corollas for the cost of 1 Ferrari engine would be better? Even though the Ferrari engine is much more powerful, it's useless without the rest of the car.
Can someone comment on the new dual die thing they’re promoting for how they make the M5 Pro and M5 Max chips?
How is that different from the silicon interposer they were using before?
The big change is the two dies don’t have to fabbed next to each other in a single wafer, which is fantastic for costs and yields. But would this affect the interconnect speed somehow?
How would the two be wired together?
Could this mean the Ultra comes back in M6 since it would be easier to fab?
They seem to market it as a technological advancement, which it is, but rather than being excited im actually worried about hidden latencies that could come with that approach.
Have you found any interesting info on that yet?
Is the M5 Max the first laptop with significantly more memory bandwidth than the M1 Max? Looks like about a 20% jump… might finally be time to re-benchmark CFD workloads.
I wonder how this compares to my M4 air with 10 GPU cores and 32 MB of RAM. My system can only run ~14B sized models at any reasonable speed. The accuracy of these sized models can be underwhelming. I am looking forward to a time when it would be nice to run models locally at a reasonable price, at a reasonable speed and with reasonable accuracy. I don't think we are there just yet.
My M3 Pro with 18gb of ram still feels like a beast. The only thing I can make it suffer with so far is generating meshes from 3D scanning, and even then I'm just patient. Apple is suffering from success with these older laptops, it's a tough sell to upgrade, even from the M1 Max folks.
I mean, they had to make them good because of the new cpu architecture, but since the emulation worked so well and overall adoption was really fast it now is a problem for them as a company. A really good problem to have though
I thought that new models were typically released in October. Have I misremembered or is this an unusual timing vs previous years? If so, I wonder why the earlier release?
I think at this point Apple will just release new versions of laptops whenever new CPU revisions and yields allow. M5 Pro wasn't ready for October so delayed until now.
I just bought a M5 Macbook Pro 2 weeks ago. Thinking of returning it and getting a M5 Pro with the same configuration but only $200 more. How should I compare M5 vs M5 Pro?
$5000 laptop you have to pay to add a power adapter… gratuitous penny pinching from Tim Cook's Apple.
It's one of those things, yes if I'm spending that much on a laptop I can afford to spend $80 on the adapter too, but does it feel good as a customer to do that or are you souring the experience of buying from you just to earn a few more dollars.
This is one thing I don't really blame Apple for, and I think everyone else will follow suit -- and not just because Apple is doing it.
The EU requires that users must be able to buy a device without a charger. It's a huge supply chain challenge to add two variants of every single SKU, one with a charger and one without. So the obvious solution is to sell the charger separately, since you need that regardless, and always sell the device without a charger. You avoid having two variants of everything that way.
Now, you could maybe argue that Apple should default to bundle a charger with your laptop, so that you'd have to uncheck a "bundle charger" checkbox on their website. But do you really care whether your laptop costs $2200 and you can buy a charger for $60 or your laptop costs $2260 and you can save $60 by removing the charger?
You can make an argument that doing it Apple's way hides a price increase. And yeah, that's probably fair. But it's not like Apple is afraid of non-hidden price increases either.
I have a huge tote box full of power bricks, most of them white Apple ones. I have a stack of 60-90W Apple USB-C ones too that I don't use cause they only have one port and are larger and worse than modern GaN units that can do 140W on one port while also pushing 30 or 60 on the others.
So, if you want one of mine, you can have one. On me. Because I'm fucking drowning in the things and appreciate not having to deal with another one.
Checking Apple's store, I can't find a cheaper configuration than $5100 for the M5 + 128GiB version.
Here in Europe, including 21% VAT, that's €6.124,00 ($7.094,35 equivalent).
Because of pricing strategies and such, the 128GiB version comes with a 2TiB SSD at minimum, and also requires the M5 Max (not Pro) at its highest configuration.
Not sure if this is new, but it should be noted that these laptops don't come with a charger any more.
70W USB-C Power Adapter (included with M5 Pro with 16-core GPU)
96W USB-C Power Adapter (included with M5 Pro with 20-core GPU, configurable with M5 Pro with 16-core GPU)
USB-C to MagSafe 3 Cable (2 m)
Devices should be offered without a charger. There's no law that states that that should be the default configuration. Nor that the charger should cost extra.
Why would you want to do that? Do you like the hardware that much, and also that much more than just an M2 (soon M3) running Asahi?
Linux in a VM would work with the usual caveats. Periphery like the built-in webcam most likely won't work. Getting codecs and DRM to run will be pain and you'll be back to use macOS for that quickly (but that's just standard pain of ARM Linux).
And your native CLI tools will continue to be from 2011 with 0 attention paid to the dev experience until it’s Swift, and we’ll continue to lock you out of running programs from other human beings we didn’t approve without a 6 step ritual in the OS. Oh and all apps will continue to constantly phone home i.e. pay for the machine so Google Adobe and Microsoft can run updaters and telemetry on it all day.
I am very excited by this, but I am a bit dampened that the maximum memory available is 128GB. I was really hoping for 256GB, which would allow me to run frontier models locally. I think with 128GB it's still feasible to use this with something like Qwen3-Coder-Next and MiniMax-M2.5, but things like Kimi-K2.5 will require significant quantization to fit and model performance will really suffer.
I'm really wanting to build proper local-first AI workflows at home, and I think Apple has an opportunity to make that possible in a way other companies aren't really focused on, but we need significantly larger memory capabilities to do it, which I know is tough in the current memory market but should be available for a cost.
With an additional $200/month subscription from Anthropic, because they noticed that the Kimi K2.5 they were able to run on their M5 comes nowhere close to matching Opus 4.6.
Local LLMs. Lots of people buy Macs due to their unified memory which obviates the need to buy a much more expensive GPU to get the same amount of VRAM.
Well that's. Just. Great. I bought a 64GB M4 Max MBP last month. I'm past the 14-day return window. I figured the M5 was near, but assumed M5 Max would come a bit later. Not sure where I came up with that.
This is always the gamble with buying a Mac. Either purchase right when the new is released, or be on the fence of your new becoming old a couple of weeks after purchase.
> Testing conducted by Apple in January 2026 using preproduction 13-inch and 15-inch MacBook Air systems with Apple M5, 10-core CPU, 10-core GPU, 32GB of unified memory, and 4TB SSD, and production 13-inch and 15-inch MacBook Air systems with Apple M4, 10-core CPU, 10-core GPU, 32GB of unified memory, and 2TB SSD. Time to first token measured with an 8K-token prompt using a 14-billion parameter model with 4-bit quantization, and LM Studio 0.4.1 (Build 1). Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Air.
Latency to the first token is not like a web page where first paint already has useful things to show. The first token is "The ", and you'll be very happy it's there in 50ms instead of 200ms... but then what you really want to know is how quickly you'll get the rest of the sentence (throughput)
Are they doubling down on local LLMs then?
I still think Apple has a huge opportunity in privacy first LLMs but so far I'm not seeing much execution. Wondering if that will change with the overhaul of Siri this spring.
I don't mind it, I open Apple stock. But I'm def not buying into their rebranding of integrated GPU under the guise of Unified Memory.
Aren't the OpenClaw enjoyers buying Mac Minis because it's the cheapest thing which runs macOS, the only platform which can programmatically interface with iMessage and other Apple ecosystem stuff? It has nothing to do with the hardware really.
Still, buying a brand new Mac Mini for that purpose seems kind of pointless when a used M1 model would achieve the same thing.
Yeah because Mac upgrade prices were already sky high, long before the component shortage. 32GB of DDR5-6000 for a PC rocketed from $100 to $500, while the cost of adding 16GB to a Mac was and still is $400.
That's likely only part of the reason. Mac Mini is now "cheap" because everyone exploded in price. RAM and SSD etc have all gone up massively. Not the mention Mac mini is easy out of the box experience.
I considered the mac mini at the time, but the mac mini only makes sense if you need the local processing power or the apple ecosystem integration. It's certainly not cheaper if you just need a small box to make API calls and do minimal local processing.
If you just need "a small box to make API calls and do minimal local processing" you an also just buy a RPI for a fraction of the price of the GMKtec G10.
All 3 serve a different purpose; just because you can buy a slower machine for less doesn't mean the price:performance of the M1 Mac Mini changes.
Do you really need Openclaw now? And not claude code + zapier or Claude code + cron?
That's the point. If you have worse CPU and GPU Windows will be sluggish (it's bloated).
https://www.youtube.com/watch?v=x4_RsUxRjKU
For the same price in API calls, you could fund AI driven development across a small team for quite a long while.
Whether that remains the case once those models are no longer subsidized, TBD. But as of today the comparison isn't even close.
And while it is stupid slow, you can run models of hard drive or swap space. You wouldn’t do it normally, but it can be done to check an answer in one model versus another.
Now extrapolating in line with how Sun servers around year 2000 cost a fortune and can be emulated by a 5$ VPS today, Apple is seeing that they can maybe grab the local LLM workloads if they act now with their integrated chip development.
But to grab that, they need developers to rely less on CUDA via Python or have other proper hardware support for those environments, and that won't happen without the hardware being there first and the machines being able to be built with enough memory (refreshing to see Apple support 128gb even if it'll probably bleed you dry).
I certainly only use Macs when being project assigned, then there are plenty of developers out there whose job has nothing to do with what Apple offers.
Also while Metal is a very cool API, I rather play with Vulkan, CUDA and DirectX, as do the large majority of game developers.
Only groups of developers more tied to Windows that I can think of are probably embedded people tied due to weird hardware SDK's and Windows Active Directory dependent enterprise people.
Outside of that almost everyone hip seems to want a Mac.
Everyone hip alright, or at least those that would dream to earn a salary big enough to afford Apple taxes.
Remember there are world regions where developers barely make 1 000 euros per month.
https://survey.stackoverflow.co/2025/technology/#1-computer-...
The US 1s? Is that why we have Deepseek and then other non-US open source LLMs catching up rapidly?
World view please. The developer community is not US only.
It wouldn’t surprise me if the deepseek people were primarily using Mac’s. Maybe Alibaba might be using PCs? I’m not sure.
Basically, too many choices to "focus on" makes non a winner except the incumbent.
People always overlook that CUDA is a polyglot ecosystem, the IDE and graphical debugging experience where one can even single step on GPU code, the libraries ecosystem.
And as of last year, NVidia has started to take Python seriously and now with cuTile based JIT, it is possible to write CUDA kernels in pure Python, not having Python generate C++ code that other tools than ingest.
They are getting ahead of Modular, with Python.
Neural Accelerators (aka NAX) accelerates matmults with tile sizes >= 32. From a very high level perspective, LLM inference has two phases: (chunked) prefill and decode. The former is matmults (GEMM) and the latter is matrix vector mults (GEMV). Neural Accelerators make the former (prefill) faster and have no impact on the latter.
Apple is in the hardware business.
They want you to buy their hardware.
People using Cloud for compute is essentially competitive to their core business.
I assume they have a moderate bet on on-device SLMs in addition to other ML models, but not much planned for LLMs, which at that scale, might be good as generalists but very poor at guaranteeing success for each specific minute tasks you want done.
In short: 8gb to store tens of very small and fast purpose-specific models is much better than a single 8gb LLM trying to do everything.
This correlation of Apple and privacy needs to rest. They have consistently proven to be otherwise - despite heavily marketing themselves as "privacy-first"
https://www.theguardian.com/technology/2019/jul/26/apple-con...
Apple's stated AI strategy is local where it can and cloud where it needs. So "doubling down"? Probably not. But it fits in their strategy.
Remains to be seen how capable it actually is. But they're certainly trying to sell the privacy aspect.
It's the best. We all turned it off. 100% privacy.
I think I'll pass on upgrading.
I also haven’t seen any improvements in the frontier models in years, and I’m anxiously awaiting local models to catch up.
"AI" (LLMs) may or may not have a bubble-pop moment, but until it does Apple get to ride it on these press releases and claims. But if the big-pop occurs, then Apple winds up with really fantastic hardware that just happens to be good at AI workloads (as well as general computing).
For example, image classification (e.g. face recognition/photo tagging), ASR+vocoders, image enhancement, OCR, et al, were popular before the current boom, and will likely remain popular after. Even if LLM usage dries up/falls out of vogue, this hardware still offers a significant user benefit.
Unified memory is a big architectural improvement.
Honestly, I think that's the move for apple. They do not seem to have any interest in creating a frontier lab/model -- why would they give the capex and how far behind they are.
But open source models (Kimi, Deepseek, Qwen) are getting better and better, and apple makes excellent hardware for local LLMs. How appealing would it be to have your own LLM that knows all your secrets and doesnt serve you ads/slop, versus OpenAI and SCam Altman having all your secrets? I would seriously consider it even if the performance was not quite there. And no need for subscription + cli tool.
I think apple is in the best position to have native AI, versus the competition which end up being edge nodes for the big 4 frontier labs.
So as most people in or adjacent to the AI space know, NVidia gatekeeps their best GPUs with the most memory by making them eye-wateringly expensive. It's a form of market segmentation. So consumer GPUs top out at 16GB (5090 currently) while the best AI GPUs (H200?) is 141GB (I just had to search)? I think the previou sgen was 80GB.
But these GPUs are north of $30k.
Now the Mac Studio tops out currently at 512GB os SHARED memory. That means you can potentially run a much larger model locally without distributing it across machines. Currently that retails at $9500 but that's relatively cheap, in comparison.
But, as it stands now, the best Apple chips have significantly lower memory bandwidth than NVidia GPUs and that really impacts tokens/second.
So I've been waiting to see if Apple will realize this and address it in the next generation of Mac Studios (and, to a lesser extend, Macbook Pros). The H200 seems to be 4.8TB/s. IIRC the 5090 is ~1.8TB/s. The best Apple is (IIRC) 819GB/s on the M3 Ultra.
Apple could really make a dent in NVidia's monopoly here if they address some of these technical limitations.
So I just checked the memory bandwidth of these new chips and it seems like the M5 is 153GB/s, M5 Pro is ~300 and M5 Max is ~600. I was hoping for higher. This isn't a big jump from the M4 generation. I suspect the new Studios will probably barely break 1TB/s. I had been hoping for higher.
5090 has 32GB, and the 4090 and 3090 both have 24GB.
So yes, the LLM should be inferencing on your prompt, but it should also be inferencing on 25,000 other things … in parallel.
Those are the compute needs.
We just need compute everywhere as fast as possible.
Do think it'll be common to see pros purchasing expensive PCs approaching £25k or more if they could run SoTA multi-modal LLMs faster & locally.
I just don't get why they're dropping the ball so much on this.
They aren’t dropping the ball, they are being smart and prudent.
Interestingly, 36-128GB models are showing as “currently unavailable” on the store page, and you can’t even place an order for them right now? But for anyone curious, it’s quoting $5099 for the 128GB RAM 14” MacBook Pro model.
No change from the previous models then, 16GB->32GB was already $400. They're cutting into their previously enormous margins to keep the prices stable, rather than hiking the prices to maintain their margins.
Isn't this it?
My M3 Pro from a few years ago for the same price had 18GB.
Interesting that this hasn't budged since the memory shortages appeared.
Apple has had enough war chests with the ability of buying the entirety of TSMC's new capacity years in advance in the past.
If I were to guess, Apple locked in their entire BOM and production capacity two years ago. That's something even the large players cannot replicate because they run cash-lean and have too many different SKUs, and the small players (Framework, System76, even Steam) are entirely left to the forces of the markets.
I'd love to have customers like Apple. Bumps $200: "it didn't change!!!"
And no power adapter included.
Try making a good product that people love?
Now it starts at $1699, a $100 bump but comes with a 1TB SSD. Previously it would have cost $1799 for the 1T SSD, so it's a $100 bump on base price but you are also getting 1TB SSD for $100 less than before.
To be fair, ever since the advent of high power USB-C PD that really, really is not needed any more, way too many power bricks are effectively e-waste.
People already have USB-C power bricks and docks everywhere and unlike pre-USB-C generations, you can use them not just across different generations of hardware, but across vendors as well.
I wonder if that would happen regardless of RAM, e.g. for tariffs etc.
> Even More Value for Upgraders
> The new 14- and 16-inch MacBook Pro with M5 Pro and M5 Max mark a major leap for pro users. There’s never been a better time for customers to upgrade from a previous generation of MacBook Pro with Apple silicon or an Intel-based Mac.
I read as "Whoops we made the M1 Macbook Pro too good, please upgrade!"
I think I will get another 2-5 years out my mine.
Apple: If you document the hardware enough for the Asahi team to deliver a polished Linux experiene, I'll buy one this year!
for example, let's say the new os depends on m5's exclusive thumbnail generator accelerator, and let's say it improves speed by a 20%.
now, your M1 notebook than on previous OSes uses standard gpu acceleration for thumbnails will not have this specialized hardware acceleration, it will have software fallback that will be 90% slower.
you won't notice it a first thought because it's stuff, fast, but it eats a bit of the processor.
multiply this by 1000 features and you have a slow machine.
I don't know how else to explain how an ipad pro cannot even scroll a menu without stuttering, it's insane how fast these things were on release
the Liquid Glass for example probably is not so great when it comes to resources. Probably works better with latest metal and hardware blocks on the GPU in M5 as opposed to using GPU cores and unified memory on 8gb M1 making latest macOS work not so great. I have the M1 8gb air and it is really slow on Tahoe. It was snappy just a couple of years ago on a fresh install.
Nothing has broken and I consistently get 4-6 hours of heavy work time while on battery. An amazing machine for the price I paid.
I still don't have a strong urge to upgrade. I could probably get by on 32GB (like my work-issued machine is) but 64GB is the right amount of headroom for me.
~9 years later, there are a lot of people still using it as their main machine, waiting until we get kicked off the corp network for lack of software support.
Incidentally, I just switched to Asahi Linux, but that was for software quality and openness reasons, rather than anything to do with performance.
Unfortunately, number always must go up (and the rate at which the number goes up, also must go up).
The new tensor cores, sorry, "Neural Accelerator" only really help with prompt preprocessing aka prefill, and not with token generation. Token generation is memory bound.
Hopefully the Ultra version (if it exists) has a bigger jump in memory bandwidth and maximum RAM.
Most stuff ends up running Metal -> GPU I thought
https://creativestrategies.com/research/m5-apple-silicon-its...
This seems even likely as the memory bandwidth hasn't increased enough for those kinds of speedups, and I guess prefill is more likely to be compute-bound (vs mem bw bound).
That's actually the biggest growth area in LLMs, it is no longer about smart, it is about context windows (usable ones, note spec-sheet hypotheticals). Smart enough is mostly solved, combating larger problems is slowly improving with every major release (but there is no ceiling).
Wondering if local LLM (for coding) is a realistic option, otherwise I wouldn't have to max out the RAM.
You mean on your first token. Whats the performance after 500 and 3000 tokens?
I genuinely don't understand why people post stuff like this. People are not informed enough to know you mean first tokens. They are going to make a mistake and buy one thinking they will get 100tk/s.
Are you working for Apple marketing? Do you have post purchase regret? I cannot imagine deliberately misleading people. Maybe you are hoping more buyers build up your ecosystem?
I think the truth is somewhere in the middle, many people don't realize just how performant (especially with MLX) some of these models have become on Mac hardware, and just how powerful the shared memory architecture they've built is, but also there is a lot of hype and misinformation on performance when compared to dedicated GPU's. It's a tradeoff between available memory and performance, but often it makes sense.
I use my laptop for development. I don't actually use most of the built in applications. My browser is Firefox, I use codex, vs code, intellij, iterm2, etc. Most of that works just fine just as it did on previous versions of the OS. I actually on purpose keep my tool chains portable as I like to have the option to switch back to Linux when I want to. I've done that a few times. I come back for the hardware, not the OS.
In my experience, if you don't like Apple's OS changes that is unfortunate but they don't seem to generally respond to a lot of the criticism. Your choices are to get further and further out of date, switch to something else, or just swallow your pride. Been there done that. Windows is a "Hell No" for me at this point. I'll take the UX, with all the pastel colors that came and went and all the other crap that got unleashed on macs over the last ten years. Definitely a case of the grass not being greener on Windows. Even with the tele tubby default desktop in XP back in the day.
I can deal with Linux (and use that on and off on one of my laptops). However, that just doesn't run that well on mac hardware. And any other hardware seems like a big downgrade to me. Both Windows and Linux are arguably a lot worse in terms of UX (or lack thereof). Linux you can tweak. And you kind of have to. But it just never adds up to consistent and delightful. Windows, well, at this point liking that is probably a form of Stockholm Syndrome. If that doesn't bother you, good for you.
So, Mac OS it is for me as everything else is worse. I've in the past deferred updates to new versions of Mac OS as well. Generally you can do that for a while but eventually it becomes annoying when things like homebrew and other development toys start assuming you run something more recent. And of course for security reasons you might just not drag your feet too long. Just my personal, pragmatic take.
Literally unusable
I already left the beta train on my iPhone because I had too many issues getting my grocery apps to allow me to place orders without going to my laptop and doing it in a web browser.
It's so bad I switched back to Chrome. I had thought Chrome had a major battery life penalty compared to Safari on Macs, but I checked more up-to-date info and apparently that's outdated.
But I think this predates Tahoe.
Also, the mix of cores have changed drastically.
- 6 "Super cores"
- 12 "Performance cores"
I'm guessing these are just renamed performance and efficiency cores from previous generations.
This is a massive change from the M4 Max:
- 12 performance cores
- 4 efficiency cores
This seems like a downgrade (in core config but may not be in actual MT) assuming super = performance and performance = efficiency cores.
I think this is a new design, with Apple having three tiers of cores now, similar to what Qualcomm has been doing for a while.
I think how it breaks down is:
- "Super" are the old "P" cores, and the top tier cores now
- "Performance" cores are a new tier and seen for the first time here, slotting between "old" P and E in performance
- "Efficiency" / "E" are still going to be around; but maybe not in desktop/Pro/Max anymore.
For example, 6 super, 8 performance, and 4 efficiency.
The base M5 has super/efficiency cores.
The Pro and Max have super/performance cores.
I believe they lower the clock speed, limit how much work is done in parallel on each core, and limit how aggressive the speculative execution is so less work is wasted.
> The industry-leading super core was first introduced as performance cores in M5, which also adopts the super core name for all M5-based products
But new "performance" is claimed to be new design (= not just overclocked efficiency core from M5?):
> M5 Pro and M5 Max also introduce an all-new performance core that is optimized to deliver greater power-efficient, multithreaded performance for pro workloads.
quotes from https://www.apple.com/newsroom/2026/03/apple-debuts-m5-pro-a...
I have not once felt the need to upgrade in years, and that’s with doing pretty demanding 3D and LLM work.
Doubt
I imagine you basically use online models exclusively, and occasionally try out local stuff.
Source: My fortune 20 company tried with M whatever, and the local llms were unusable.
The high memory Macs have been great for being able to run LLMs, but the prompt processing has always been on the slow side. The new AI acceleration in these should help with that.
There are also workloads like compiling code where I’ll take all the extra speed I can get. Every little bit of reduced cycle time helps me finish earlier in the day.
And then there’s gaming. I don’t game much, but the M1 and M2 era Apple Silicon feels sluggish relative to what I have on the nVidia side.
M5 Max maxes out at 128GB, so that will have to wait for the eventual M5 Ultra anyways.
Even if a new device is a small upgrade from last year's model, it can be a giant upgrade for other people.
CoreImage - GPU accelerated image processing out of the box;
ML/GPU frameworks - you can get built-in, on device's GPU running ML algorithms or do computations on GPU;
Accelerate - CPU vector computations;
Doing such things probably will force you to have platform specific implementations anyway. Though as you said - makes sense only in some niches.
I think I read somewhere long time ago that Capture One is also using Qt for GUI, though cannot find this anymore, so probably not true.
Now it starts at $1699, a $100 bump but comes with a 1TB SSD. Previously it would have cost $1799 for the 1T SSD, so it's a $100 bump on base price but you are also getting 1TB SDD for $100 less than before.
For example, up until MacBookPro M2, MacBookPro M2 came with M2 Pro chip.
However, starting with M3, Apple lowered the MacBookPro MSRP to $1599, but its base configuration was downgraded to M3 chip from M3 Pro. To get the M3 Pro, you had to pay $1999. There's substantial performance between the two.
Same with M4. To get the M4 Pro chip, you had to pay $1999.
Now to get M5 Pro chip, it's $2199. Still a good value, but just saying it's a deviation from the trend.
It's the first time I've ever been so repulsed by a design that I actively avoid it just... out of sheer preference.
The prompt processing sped up.
Not the output generation.
M4 was notoriously slow at this compared to DGX etc.
I have a Intel-based 2019 Macbook Pro still and I have NEVER in its lifetime gotten even half of what they are claiming here. These days if I run it from battery I might get 90 mins.
That said I had a maxed out Macbook Pro M4 Max on order but just cancelled it right now and will get this new M5 Max one for basically the same price. Once I saw that they didn't up the price of memory (I don't know how it doesn't affect them) I canceled my order.
https://github.com/Sikarugir-App/Sikarugir
For those of us with astigmatism it's really night and day experience.
This is the important statement. 614GB/s is quite decent, however a NVIDIA RTX 5090 already offers 1,792 GB/s (roughly 3x) of memory bandwidth, for comparison.
You can buy two m5 pro base model for the same price as a single 5090...
In Europe I can get a 128gb mac studio m4 max for 300 euros more than a 5090 (for which you still need to buy a power supply, motherboard, cpu , &c.)
How is that different from the silicon interposer they were using before?
The big change is the two dies don’t have to fabbed next to each other in a single wafer, which is fantastic for costs and yields. But would this affect the interconnect speed somehow?
How would the two be wired together?
Could this mean the Ultra comes back in M6 since it would be easier to fab?
They seem to market it as a technological advancement, which it is, but rather than being excited im actually worried about hidden latencies that could come with that approach. Have you found any interesting info on that yet?
I think at this point Apple will just release new versions of laptops whenever new CPU revisions and yields allow. M5 Pro wasn't ready for October so delayed until now.
They also probably had RAM contracts in place far enough in advance to avoid the worst of the price spikes.
And another rumor said these are going to be updated again this fall but I’m not sure about that. With OLED screens and M6 (supposedly).
Wish it was Blender though ;)
It's one of those things, yes if I'm spending that much on a laptop I can afford to spend $80 on the adapter too, but does it feel good as a customer to do that or are you souring the experience of buying from you just to earn a few more dollars.
https://appleinsider.com/articles/25/10/15/eu-gets-what-it-a...
In the US they provide one in the box free of charge.
The EU requires that users must be able to buy a device without a charger. It's a huge supply chain challenge to add two variants of every single SKU, one with a charger and one without. So the obvious solution is to sell the charger separately, since you need that regardless, and always sell the device without a charger. You avoid having two variants of everything that way.
Now, you could maybe argue that Apple should default to bundle a charger with your laptop, so that you'd have to uncheck a "bundle charger" checkbox on their website. But do you really care whether your laptop costs $2200 and you can buy a charger for $60 or your laptop costs $2260 and you can save $60 by removing the charger?
You can make an argument that doing it Apple's way hides a price increase. And yeah, that's probably fair. But it's not like Apple is afraid of non-hidden price increases either.
So, if you want one of mine, you can have one. On me. Because I'm fucking drowning in the things and appreciate not having to deal with another one.
Here in Europe, including 21% VAT, that's €6.124,00 ($7.094,35 equivalent).
Because of pricing strategies and such, the 128GiB version comes with a 2TiB SSD at minimum, and also requires the M5 Max (not Pro) at its highest configuration.
Not sure if this is new, but it should be noted that these laptops don't come with a charger any more.
Devices should be offered without a charger. There's no law that states that that should be the default configuration. Nor that the charger should cost extra.
Linux in a VM would work with the usual caveats. Periphery like the built-in webcam most likely won't work. Getting codecs and DRM to run will be pain and you'll be back to use macOS for that quickly (but that's just standard pain of ARM Linux).
I'm really wanting to build proper local-first AI workflows at home, and I think Apple has an opportunity to make that possible in a way other companies aren't really focused on, but we need significantly larger memory capabilities to do it, which I know is tough in the current memory market but should be available for a cost.
128 GB maximum.
Sigh.
LOL. is it repairable? probably not.
You can run open source models like Kimi K or Qwen locally. Apple recently updated Xcode 26.3 to support local models.