Apple’s base Mac Mini went from $599 to $899 on August 25, a 50% jump for what used to be the company’s cheapest way into the Mac ecosystem. Most of the coverage that followed walked through the new M6 chip’s specs — a 12-core CPU, a 2nm manufacturing process, a redesigned Neural Engine — as if that alone explained the number. It doesn’t. AppleInsider’s own headline called it “a RAM and SSD constrained environment,” and that phrase does more explanatory work than any spec sheet: the Mac Mini’s price didn’t jump because the chip got that much better. It jumped because building a computer got more expensive, for reasons that have little to do with Apple’s engineering.
What Changed in the Box
The outgoing base Mac Mini shipped with an M4 chip, 16GB of RAM, and 256GB of storage for $599 (Apple had briefly sold a 512GB M4 configuration for $799 before discontinuing the $599 tier entirely). The new base model has an M6 chip, still 16GB of RAM, still 256GB of storage, for $899. Apple’s own performance claims put the M6 at roughly 4x the M4’s AI throughput, with storage and graphics performance roughly doubled — genuine generational gains, especially for anyone running local AI models, which has become a real use case for this specific machine rather than a hypothetical one. Enthusiasts have reportedly been paying inflated resale prices for M4 Mac Minis just to get a cheap local-inference box, which tells you the demand pressure was already there before this price even landed.
But notice what didn’t change: RAM and storage are identical to the outgoing model. You’re not paying $300 more for more memory or more disk. You’re paying it for a new chip generation and whatever it now costs Apple to source the same 16GB/256GB configuration it sourced a year ago.
The Memory Market Explains the Gap the Spec Sheet Doesn’t
This site covered the DRAM and NAND price surge driving up laptop and TV prices a few weeks ago: AI datacenter demand pulling wafer capacity away from consumer memory, DRAM contract prices up roughly 90% in a single quarter, a single AI server needing eight to ten times the DRAM of a traditional one. That same mechanism sits underneath the Mac Mini’s price tag, even though Apple never says so directly in its own materials. A 2nm process node and a new Neural Engine are real, chargeable upgrades, but they don’t typically double a base configuration’s price gap in a single generation on their own. The blunt “RAM and SSD constrained environment” framing from AppleInsider’s coverage is the tell: hardware press covering Apple specifically is now reaching for the same explanation used for PS5 price hikes, Steam Deck OLED’s 43-46% jump, and TV panel pricing. This isn’t purely an Apple story. It’s a memory-market story wearing an Apple product.
The Mac Studio Tells the Same Story, Louder
Apple didn’t just refresh the Mini. The Mac Studio moved to M5 Max ($2,499 starting, 36GB unified memory, 512GB storage) and M5 Ultra ($5,499 starting, 96GB unified memory, 1TB storage) configurations. Those numbers matter here because unified memory is exactly the resource getting squeezed industry-wide — Apple’s own chips share memory between CPU and GPU rather than using separate RAM and VRAM pools, which means Apple is buying into the same tightening DRAM supply as everyone else, just packaged differently. A $5,499 starting price on a desktop is a genuinely high number even by Mac Studio standards, and it arrived the same day as the Mini’s jump. That’s a pattern, not a coincidence.
What Apple Isn’t Saying Out Loud
Apple’s own product pages frame both machines purely around performance: faster Neural Engine, faster storage, more graphics cores. Nowhere in Apple’s official materials is there an acknowledgment of component costs driving price, which is standard practice for a company that prefers to sell price increases as value increases rather than as pass-through costs. That’s worth knowing specifically because it shapes how the price should be read: a shopper comparing “what did I get for $300 more” against Apple’s own performance claims is comparing against a number Apple has every incentive to make look as generous as possible. Reading the AppleInsider and industry-analyst framing alongside Apple’s own materials gives a fuller picture than either alone — the performance gains are real, but they’re being used to make a cost increase that would exist either way look like a value trade instead.
Who Should Buy the New Mac Mini, and Who Shouldn’t
If you’re using a Mac Mini as a general-purpose desktop — browsing, office work, light photo editing — the M6 generation is a real but modest upgrade over M4, and you’re now paying nearly $300 more to get it. That’s a worse deal in almost every practical sense than it was a year ago, and there’s no “wait for a sale” option here: Apple doesn’t discount new hardware in its first year, and the memory-market pressure pushing this price up isn’t expected to ease soon. Analysts covering the broader DRAM shortage have pointed to 2027 as the earliest realistic relief window, not next quarter.
If you’re specifically buying this machine to run local AI models or agent workloads, the calculus is different. The Neural Engine redesign and the roughly 4x AI throughput claim are the reason this specific machine has developed a following beyond Apple’s usual desktop buyers, and for that narrow use case, the extra cost buys something the M4 generation genuinely couldn’t do as well. That’s a smaller audience than “anyone shopping for a budget Mac,” but it’s the group this price increase is arguably aimed at holding onto rather than losing to a cheaper Windows mini-PC or a cloud AI subscription instead.
What to Check Before You Buy
A few things worth confirming before you order, since the base configuration is a worse value than it’s been in years:
- Whether 16GB of unified memory is enough for what you’re doing. It was borderline for the outgoing M4 model at $599; at $899 it’s a much harder sell to not upgrade, and RAM on Apple Silicon can’t be added after purchase. You’re locked into whatever tier you buy on day one, permanently.
- Whether the M5 Pro tier ($1,699, 24GB RAM, 512GB storage) makes more sense for your budget than stretching for the base M6 model and immediately feeling storage-constrained. The jump between tiers is steep, but so is the jump from 256GB to anything usable for serious local AI work, where model files alone can run tens of gigabytes.
- Whether a refurbished or open-box M4 Mac Mini, now discontinued, is sitting on a retailer’s clearance shelf. Apple’s own store won’t sell you the outgoing model anymore, but third-party retailers frequently clear out the previous generation at a real discount once a new one ships. A direct price check across a few retailers before assuming the M6 is your only current option can genuinely save several hundred dollars for a buyer who doesn’t need the new Neural Engine specifically.
- Whether your workload is CPU/GPU-bound or Neural-Engine-bound. The 4x AI throughput figure only matters if what you’re doing routes through the Neural Engine at all; standard office and browser workloads barely touch it, and for that use case the M4-generation performance you’re being asked to pay $300 more to leave behind was already more than sufficient.
The new Mac Mini ships September 22. If your use case doesn’t specifically benefit from the new Neural Engine, there’s little upside to buying on launch day rather than checking whether last year’s model is still findable at its old price somewhere first.
