The stuff Hacker News actually recommends.

Laptops · Apple MacBook Pro

Apple MacBook Pro M1 Max

142owners with an opinion
77%of verdicts positive
80/27positive/negative since 2024
7 yrslongest happy owner
8happy after 3+ years

Praised for performance and reliability; criticized for battery and display.

What owners wrote

Likes itOwned 7 yearsStill going
I'm a lunix engineer, probably 25+ years of lunix now (I started on bsd, thats why I say lunix), I've worked on macbook pros for the last 15 years. I had to reinstall macos for the FIRST TIME ever this year because their current version Tahoe is absolute trash that slowed down my M1 Max so my much I finally gave up and wiped it. I also won't go near an x86 processor ever again after using the mX/arm procs. We're paying for the hardware. The 100x better screens, speakers, battery life, the apps on macos blow linux apps away even if all of them, even the most simple apps, cost $20+. But with that said Apple has also priced me out of buying another MBP for a long time with how expensive their 64gb+ versions are. My m1 max was $3500 and I can't believe I paid that, but here I am 7 years later still using it and it cranks. I'd have replaced 3 x86 laptops by now. Oh, and I had the entire laptop replaced for $99 when its battery got bad and the screen cracked last year thanks to applecare, though that did require me paying for applecare at $99/yr each year. But that still doenst add up to another $3500 laptop. I just drove to the apple store dropped it off and picked it up a week later.
Recommends itOwned 5 yearsStill going
> And nobody in the right mind would buy a used Macbook for the same price, just because it's more powerful. I would buy a used Macbook Pro for the same price, just because it is more powerful. I have bought used Macbook Pros for many years, and have never had any problem. Apple gear is durable, and I've never had any SSD issues or problems with dead pixels. The laptop I'm typing this on is a 2021 M1 Max, so 5 years old now, and I just did an SSD health check, it is 100% after 5 years of heavy use. It easily has 5 more years of life.
Likes itOwned 4 yearsStill going
I'm still running a 4 year old M1 Max 32GB Mac Studio as my daily and really haven't found anything that it shows sign of slowness/age with. That said I don't do any local AI or 8K video editing type stuff. On upgrade I could probably live on a moderate/high spec Mac mini since the only reason I went Studio originally was the resolution/number of displays it could support.
Likes itOwned 4 years
Macs on the other hand are extremely stable. In my 4 years of using my MacBook Pro M1 Max, I’ve only restarted during OS updates. There were maybe a handful instances where it froze and I forced restart. Other than that, I only put it to sleep every time and it works like a charm. I use it for heavy duty software development and experimentation with local models, so it’s even more surprising!
Likes itOwned 4 yearsStill going
Another innovation is their line of M-series chips. I'm typing this on an M1 Max and this is by far the best combination of industrial design and physical hardware that I have used over the past 20 years. I've had it for four years, three new generations of chips have come out since then and this thing chews through anything I throw at it. A ground-breaking, industry-changing innovation like the iPhone is like lightning in a bottle. It would be insane to think Apple can capture lightning every 5 years like clockwork.
Likes itOwned 4 yearsStill going
The jump from Intel to M* is huge. Whether that's worth it for you - hard to say. You can get a good deal on a refurbished or used M* MBP and try it out. My 2021 M1 Max MBP is still going strong; so strong I just can't justify a new one. Biggest thing to note is how many external displays you want to drive. I got the M1 Max to drive my 2-4.
Likes itOwned 3 years
I have a three year old M1 Max, 32gb RAM. Llama 8bn runs at 25 tokens/sec, that’s fast enough, and covers 80% of what I need. On my ryzen 5600h machine, I get about 10 tokens/second, which is slow enough to be annoying. If I get stuck on a problem, switch to chat gpt or phind.com and see what that gives. Sometimes, it’s not the LLM that helps, but changing the context and rewriting the question. However I cannot use the online providers for anything remotely sensitive, which is more often than you might think. Local LLMs are the future, it’s like having your own private Google running locally.
Likes itOwned 3 yearsStill going
For many years, I was one of Apple's highest-end customers. I needed the fastest machines they offered for the workloads I threw at them. I made music in Logic Pro with a zillion plugin instruments and effects and I ran a ton of apps with a ton of open windows when I wasn't doing that. I always bought the PowerMacs and then Mac Pro and eventually the iMac Pro when Apple painted themselves into a thermal corner with the Mac Pro. I bought a MacBook Pro M1 Max in 2021 and started using it as my primary machine driving external Studio Displays. It's the first computer I've ever owned that doesn't feel the least bit slow as I approach the three year mark. I've been asking myself when I think I'll upgrade and I truly don't know; probably not before some incredible new software feature isn't available on this hardware. I used to replace machines when the three year AppleCare coverage expired or soon after. I can't see that happening any longer. To some degree, this isn't even about Apple Silicon (though that sure helps). Even my iMac Pro with Xeon processor was approaching a point where I'd have got to this milestone eventually. Apple Silicon accelerated the jump by a good margin. I have no doubt I'd be able to tell the difference if I got an M4 laptop whenever they arrive, but I'd struggle to justify it to myself. (I should mention, in fairness, that I've been on a hiatus from making music since partway through the pan and that undoubtedly makes a difference. But even then, I think
Likes itOwned 2 years
As mundane and common as the information I’m about to share may be, it’s the reality of MacBook ownership for many - dare I say most? I don’t buy new MacBooks. I get them refurbished from MicroCenter, Apple, or elsewhere. In 2024, I picked up a 2021-released 14” M1 Max MacBook Pro w/ 2 TB and 64 GB for $1800 from Microcenter. Because Apple is so hostile to repairs, I have been paying for AppleCare+ and will continue for some time. But I don’t see myself truly needing to replace this machine for another 2-3 years. It’s my daily driver for BBedit, some IDEs, music apps, Firefox with dozens of tabs, Mail, Messages, iTerm2 with half a dozen tmux sessions. True - my computing needs are not demanding. But for those of us who offload commuting to remote servers and mostly use thin clients, the computer is a joy to use. I’ve been dabbling in Ableton Live Suite 12 Lite, which came with my Native Instruments Komplete Kontrol S88 MKII midi keyboard, and this machine never breaks a sweat. The M-series are seriously as close to BIFL as a consumer laptop can be. Meanwhile, my wife’s daily driver is a 2013 13” MacBook Air w/ 128 GB SSD and 8 GB RAM. And an Intel i7 from that era. I won’t defend this though. She’s insane, and I don’t know how she lives with herself. I purchased it used from a Craigslist seller to be my daily driver 14 years ago, and it lasted until I upgraded to a 2020 M1 MacBook Air base model, which I later donated to a friend in need.
Likes itOwned 2 yearsStill going
The M1 Max that I have is easily the greatest laptop I've ever owned. It is fast and handles everything I've ever thrown at it (I got 32 GB RAM), it never, ever gets hot, I've never heard a fan in 2+ years (maybe a very soft fan if you put your ear next to it). And the battery life is so incredible that I often use it unplugged. It's just been a no-compromise machine. And I was thinking of upgrading to an M3 but will probably upgrade to an M4 instead at the end of this year when the M4 maxes come out.
Neutral
I have subscriptions for necessary work and even pay for api calls. The machine with the 3090 I built years ago and suddenly it’s worth 4x plus what I paid. It is for testing, but I use the model I run on it and the model I run on an mbp M1 Max with 64gig as workers for specifically scoped work. I’d love to have a few dgx sparcs or the like to do heavy local stuff, but that money is best paying the pimps for now.
Doesn’t like it
Apple never drops prices, never. It will go straight into their margin. I have an M1 Max but I am not paying 50% more than a few months ago to upgrade. I have bought Apple laptops since 2004 but I can’t justify the current extortionate pricing, so I’ll sit it out (permanently if need be).
Doesn’t like it
I just bought a portable keyboard for my iPhone - coupled with a SSH client, its quite a useful combo these days. I had to prove the OP wrong though; my M1 Max's keyboard died this year, and I had to upgrade to the M5 max this season, after the hikes. I kissed 9k EUR goodbye for something not much better functionally than the M1 Max it replaced, bought for 3.4k Canadian dollars in the wonderful 2021 (that was barely over 3200 EUR back then).
Likes it
Here are the numbers that got me buying a used M1 Max for LLM work. M1 Max 400GB/s M4 Pro 273GB/s M5 Pro 307GB/s Only Max are worthwhile for LLMs. An Ultra is just two Mac chips on a die, but I think they're just used in Mac Studio. I'll check it that inference engine! Thanks! Edit: it doesn't give numbers for unsloth/Qwen-3.6-35B-A3B 4bit, but I'm using mlx-serve right now and getting petty good numbers (and it took a while to tweak everything properly!)
Likes it
> "Man you Apple Fans really live under a rock... ...refurbished 3090 is 2500" Before you make assumptions, there are valid reasons for the choice I made: I paid $1k for a refurbished M1 Max 64GB macbook that can run Qwen 3.8 27B at 24tok/sec (looks like they are about $1.5k on ebay). In my case, Apple was a lot cheaper (used) if you're looking at price-per-token. Yes, it's almost twice as slow as a 3090, but it's also half the price /performance ratio. Qwen 3.8 27B Q4 with 64k: 3090: ~40tok/sec M1 Max : ~24tok/sec (MLX) In fact, the Max can even do 265k context which isn't possible with a single 3090 last I checked (without dropping to Q3-Q2)
Neutral
Shouldn’t be embarrassing to admit. It’s a huge new subject area and it’s completely exhausting to learn. I still feel like a total novice. I have an M1 Max (more memory bandwidth) so I am not sure how much advice I can give about models, but I would suggest maybe trying the new LFM2.5-24B-A2B. It’s surprisingly good. If you want to mess about with multimodal work, Gemma 4 12B is really impressive. (I mostly use Muse Glimmer but it might be rather slow on an M1 Pro) If you want to have the space to learn the terminology, tinker with what you can control etc., then you have three easy options: LM Studio, Unsloth Desktop or the llama.cpp native Mac app. But don’t get put off by the llama.cpp command line; most of these tools are running it behind the scenes, and most entries on Hugging Face have recommended parameters. MLX is an option but so far I haven’t found it to have an edge over simple GGUF on Metal. I have explored running OpenCode inside a VM (you can “remote into it” from the OpenCode GUi desktop app). You might want to look at Paseo, which can do a similar thing but with many more coding agents. I am quite intrigued by Osaurus; on macOS 26 and up it looks like it will manage a container for you. You might also want to look at Orchard. Not done this myself yet but I can see the appeal. https://orchard.andon.dev/ai.html I am an experienced web dev in my fifties. Six months ago everything I have written above would have been total gibberish to me. Learning some of the…
Likes it
I’ve felt the same way since way back then ever since reading about this guy’s setup[0], so I paired a 11” MBA with the quad core Mac Mini 2012. I’ve kept that philosophy over the years, the MBA was replaced by the 2017 12” retina MacBook and then again with the 13” M2 MBA. The Mini gave way to the 2018 Mini, then a M1 Max Studio. I even attempted replacing the MacBook with an iPad though that didn’t really workout. It’s the best of both worlds really, enough power to do everything when on the go and even more power back at the office. 0: https://elliotjaystocks.com/blog/a-note-about-my-new-mac-set...
Doesn’t like it
I so wanted to be able to use Qwen3.8-27B locally on my fully loaded M1 Max 64GB when spider-mario recommended I use MTPLX[1], but I still found the general MLX MTP acceleration to be so poor, I'd rather just use cheap tokens from OpenCode Zen and Go. It's just too slow of a model. I know, I know there's new hardware, but my business paid like 4-5 grand for this MBP at the time, and I just don't feel the need to pay 7-8 grand to step up to current hardware when token spend is what it is. [1]: https://news.ycombinator.com/item?id=49611128#49612229
Doesn’t like it
I just looked and I think what I'm talking about, where they folded more specific permission callouts under a broad term "Full disk access" "local devices" etc used to be more granular before they updated that UI to match the iphone. If I got a jumpscare permission prompt on macos, I used to actually be concerned. Now they're all jumpscares for basic stuff. But I could be misremembering. I haven't liked any of the recent macos releases. I will begrudgingly install Golden Gate to hopefully fix my m1 max 64gb slowing ot a crawl with Tahoe.
Doesn’t like it
I have a 16" M1 Max that I only got because it was $1500 cheaper than MSRP, and it sucks on planes. I have really long arms and I can barely get it out of my bag without elbowing my neighbor. A few years ago I saw some very interesting custom ergonomic setups optimized for traveling + flying. One person with a thinkpad is able to get the monitor to be 180 degrees flat w/ the keyboard, and can hang it off the seat. He also brings a split ergo keyboard with a lap mount. Another person did something similar with a M1 laptop, but needs an Ipad to act as the external monitor (laptop stays in bag) with a built and designed from scratch split ergo keyboard.
Doesn’t like it
I just ran Unsloth's Q4_K_M quant on Macbook Pro M1 Max (2021), 32 GB RAM. Getting about 9 tok/s. Reported RAM usage 15.9 GB. It's okay, but a bit slow for my taste. An example of a query and a response is below. It was done with 9.09 tok/s, 893 tokens, 7.56 seconds to first token, thought for 1 minute and 9 seconds. Question: I gave the model this truth table and asked to tell me what boolean function would produce the output given A, B as inputs. | A | B | Output | | :--- | --- | ---: | | (T) | (F) | (T) | | (T) | (T) | (F) | | (F) | (F) | (F) | | (F) | (T) | (F) | Answer: (note this is part of a larger conversation, where I was getting it to help me with some boolean logic): The boolean condition for your table is simply: ```rust A && !B ``` ### Why? Your table shows the output is `true` **only** when `A` is true and `B` is false. That's exactly what `A && !B` expresses. | A | B | `A && !B` | Output | |---|---|-----------|--------| | (T) | (F) | `T && T = T` | | | (T) | (T) | `T && F = F` | | | (F) | (F) | `F && T = F` | | | (F) | (T) | `F && F = F` | | ### In Rust: ```rust if A && !B { // output is true } else { // output is false } ``` This is the most direct and idiomatic way to express that logic. Let me know if you need it adapted back to your `live_ticks` / `on_disconnect` context!
Doesn’t like it
At least for the Gemma4-26B-A4B, Token-gen speed with OMLX is far worse on my M1 Max 64GB Macbook, compared to llama-server: Quick benchmark on M1 Max 64GB, Gemma 4 26B-A4B (MoE), comparing matched dynamic 4-bit quants. Workload was Claude Code, which sends ~35K tokens of input context per request (system prompt + tools + user message): llama.cpp (unsloth/gemma-4-26B-A4B-it-GGUF:UD-Q4_K_XL, llama-server -fa on -c 131072 --jinja --temp 1.0 --top-p 0.95 --top-k 64): - pp ≈ 395 tok/s - tg ≈ 40 tok/s oMLX (unsloth/gemma-4-26b-a4b-it-UD-MLX-4bit, omlx serve --model-dir ~/models/omlx, with sampling.max_context_window and max_tokens bumped to 131072 in ~/.omlx/settings.json): - pp ≈ 350 tok/s - tg ≈ 5–13 tok/s Same model family and quant tier. Prompt processing is comparable, but oMLX's token generation is 3–7x slower than llama.cpp's Metal backend. Counter-intuitive given MLX is Apple's native ML framework.
Doesn’t like it
> Noone is going to develop on this after all. Because it doesn't have twice the ram. Otherwise it was a no brainer complementary machine, especially for users like me that work primarily on desktop and don't want to bring the much heavier macbook pro around. I've got both the m1 max and m3 max (16") and I absolutely hate carrying them around yet I have to, because even on vacations I may have to log and fix a bug in prod blocking the company so to me, weight is absolutely a primary factor for a notebook, and this would've been perfect at just twice the ram.