Apple Mac Studio

Long-term owners

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

Oct 2026 · comment

What hackers pick instead

• Apple MacRecommended for excellent build quality and tight hardware-software integration that makes it reliable and easy to use, though critics dislike macOS UI changes and pricey upgrades.453recommend
• Apple M1 MacStill fast enough for web dev, everyday work and Linux years after release, and cheap used. Critics note soldered, hard-to-upgrade storage.48recommend
• Apple IIRecommended as an open, expandable, hacker-friendly 1977 computer that was everywhere in classrooms; critics cite its price and limited graphics and hardware.46recommend
• Apple MacintoshRecommended for its graphical interface, attention to detail, and strength in visual and creative work. Critics say it started out costly and limited, with weak multitasking and slow performance for demanding tasks.34recommend
• Apple M-series MacPeople recommend M-series Macs for battery life, displays, trackpads, quiet operation and efficient chips. Common caveat: poor Linux support and limited hardware documentation.34recommend

Every quote

The 40 most detailed of 649 comments. Each links to the original.

1. 1TB fits a frontier-ish model in memory without massive quantization, which is a very interesting capability for a non-rack piece of compute.

   Recommends · Aug 2026 · in “New Mac Studio with M5 Max and M5 Ultra”

2. 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

   Recommends · Oct 2026 · in “New 14‑inch MacBook Pro with M6”

3. recent releases of the quantized GLM 5.2 models are actually pretty speedy and are probably as good or better than any online model

   Recommends · Jun 2026 · in “Apple increases MacBook and iPad prices by 20%”

4. The M3 Ultra's prefill is super slow with really large models, 3-5 minutes while it digests the new additions to its context

   Mixed · Apr 2026 · in “How good is Mac Studio M3 Ultra for Trillion param models like DeepSeekv4?”

5. an M3 Studio Ultra outpaces the M4 and ties the M5 on single work at a time, outpaces both doing multiple work at a time.

   Recommends · Jun 2026 · in “Running local models is good now”

6. This unified architecture makes Apple hardware very good for AI work, where latency between RAM and CPU is very important

   Recommends · Aug 2026 · in “Apple caught off guard by AI demand for Mac Mini and Mac Studio”

7. I suppose the Mac Studio route is a good option here, I think Apple make the most energy efficient high-memory options.

   Recommends · Jan 2026 · in “Show HN: Replacing my OS process scheduler with an LLM”

8. people are successfully and efficiently running 4-bit quantized GLM 5.2 on MacStudio 512GB M3 Ultras at over 60 tok/s

   Recommends · Jun 2026 · in “Apple increases MacBook and iPad prices by 20%”

9. it'll be slow: even with the proper multi-node inference setup, the thunderbolt link will be a comparative bottleneck

   Mixed · Jan 2026 · in “Kimi Released Kimi K2.5, Open-Source Visual SOTA-Agentic Model”

10. I've been able to run local, quantized LLMs like Qwen 3.6 using oMLX + Opencode pretty comfortably and productively.

    Recommends · Jun 2026 · in “Unsloth GLM-5.2 – How to Run Locally”

11. which I expect would be pretty amazing as far as quality/recall. The only downside is that the speed is a lot slower

    Mixed · Aug 2026 · in “Apple caught off guard by AI demand for Mac Mini and Mac Studio”

12. Because you can run Qwen 3.8 Flash Next, Laguna S 2.1 and other medium-sized models that simply don't fit on a 5090?

    Recommends · Sep 2026 · in “M5 Ultra Mac Studio Review”

13. it is still faster than the rest of the line and the only self-contained computer that can handle models that large

    Recommends · Mar 2026 · in “Apple's 512GB Mac Studio vanishes, a quiet acknowledgment of the RAM shortage”

14. For local dev just being able to fit the weights in memory is often way more important than token generation speed

    Recommends · Jul 2026 · in “Apple Will 'Watch Everything Burn' When the AI Bubble Bursts”

15. A Mac Studio is a much better buy in terms of memory bandwidth, but impossible to buy in a 128 GB configuration

    Mixed · Jul 2026 · in “AMD Ryzen AI Halo – $4k AI Dev Kit”

16. IMO getting a studio for LLMs is a wiser choice since it doesn't heat up and there are no batteries to degrade.

    Recommends · Oct 2026 · in “MacBook Pro with PROLED XDR Display and Touch”

17. I genuinely can see the value if you want to host something internally for sensitive and important information

    Recommends · Dec 2025 · in “Big GPUs don't need big PCs”

18. it's much faster, both subjectively and objectively by looking at how long it takes to run the full test suite

    Recommends · Aug 2026 · in “Apple introduces M6 and M5 Ultra”

19. uses an order of magnitude -less- energy than the computer it replaced, and is faster in almost every aspect.

    Recommends · Feb 2026 · in “Ask HN: Anyone Using a Mac Studio for Local AI/LLM?”

20. this machine you can haul in an Ice Road Truck to the North Pole and do inference there in an off grid shack

    Recommends · Aug 2026 · in “New Mac Studio with M5 Max and M5 Ultra”

21. Why people buy the Studio with the high ram config is actually the unified memory. This is unique to Apple.

    Recommends · Mar 2026 · in “Apple's 512GB Mac Studio vanishes, a quiet acknowledgment of the RAM shortage”

22. It's already clear to me that M5 Mac Studio is more cost-effective than anything you can run on open router

    Recommends · Sep 2026 · in “M5 Ultra Mac Studio Review”

23. it is inferior to nvidia precisely during prefetch (processing prompt/context) before it really gets going

    Mixed · Feb 2026 · in “GLM5 Released on Z.ai Platform”

24. The people buying Apple Silicon for inference general aim for the Mac Studios with enormous amounts of RAM

    Recommends · May 2026 · in “Was my $48K GPU server worth it?”

25. It can notably run some of the best open weight models with little power and without triggering its fan.

    Recommends · Feb 2026 · in “Qwen3.5: Towards Native Multimodal Agents”

26. 256 is the sweet spot, you can run large open weight models at decent speeds for full private inference.

    Recommends · Sep 2026 · in “M5 Ultra Mac Studio Review”

27. The GFLOPS and memory bandwidth just arne't there even though it can hold a much larger model in memory.

    Mixed · May 2026 · in “Was my $48K GPU server worth it?”

28. the Mac isn’t that slow and can use much better model compared to a similarly priced Nvidia workstation

    Mixed · Aug 2026 · in “New Mac Studio with M5 Max and M5 Ultra”

29. Mac Studios don't share unified memory across machines, Thunderbolt isn't remotely comparable to NVLink for expert exchange, and bandwidth becomes the bottleneck immediately.

    Criticizes · Jan 2026 · in “Kimi Released Kimi K2.5, Open-Source Visual SOTA-Agentic Model”

30. Regret on a $10k desktop rendered obsolete for purpose (the 512GB of RAM only has so many applications) months later is not a great look.

    Criticizes · Mar 2026 · in “Apple's 512GB Mac Studio vanishes, a quiet acknowledgment of the RAM shortage”

31. Your Mac Studio cannot run frontier LLMs at an interactive speed, even Apple has given up on using it as an inference backbone.

    Criticizes · Jul 2026 · in “Apple Silicon Exec Explains Mac Mini AI Demand and On-Device Future”

32. A single Mac Studio w/ M5 Ultra and half that RAM (256 GB) is around $11k, and the 512 GB option isn't even available yet.

    Criticizes · Oct 2026 · in “Muse Spark Open”

33. The performance with large context would have been unusable especially if you need to serve more then a single person.

    Criticizes · Apr 2026 · in “Kimi K2.6: Advancing Open-Source Coding”

34. they think those gigabytes of memory on Mac studio/mini is a lot only to find out that its "meh" on context of LLMs

    Criticizes · Mar 2026 · in “Apple Introduces MacBook Pro with All‑New M5 Pro and M5 Max”

35. 8000 dollars for a 16TB harddisk? At that price point I can load out two full PC's for the price of a single disk

    Criticizes · Aug 2026 · in “New Mac Studio with M5 Max and M5 Ultra”

36. The Mini and especially the Studio form factor will have better cooling, lower prices for comparable specs and a much higher ceiling in performance and memory capacity.

    Recommends · Feb 2026 · in “Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers”

37. Apples ProRes codec is only licensed to run in high quality mode on a Mac, and so my Nvidia PC can't do what I need.

    Recommends · Aug 2026 · in “Apple caught off guard by AI demand for Mac Mini and Mac Studio”

38. Apple computers tend to have excellent resale value, and Mac Minis/Studios have the least depreciation of them all.

    Mixed · Aug 2026 · in “Apple introduces M6 and M5 Ultra”

39. Apple arguably are making AI workstations that can do inference and training by their Mac Studio and Macbook Pros

    Recommends · May 2026 · in “Apple, Intel have reached preliminary chip-making deal”

40. we should all be buying a fully loaded Mac Studio (128GB of ram, 20 CPU cores, a lot of GPU and Neural cores.)

    Recommends · Jul 2026 · in “Apple Silicon Exec Explains Mac Mini AI Demand and On-Device Future”