Mac mini M6 is the first 2-nanometer chip in Apple’s history. On August 25, 2026, the company released an updated Mac mini with this exact chip. Alongside it, an updated Mac Studio with the M5 Ultra chip was also released. We break down what makes these new products special specifically for running AI models locally.
Mac mini M6: What’s New
The M6 chip received a 12-core CPU and a 12-core GPU — an increase compared with the 10-core blocks in the M4. The main new feature is Apple’s first dual Neural Engine with two 16-core blocks working simultaneously, instead of one in previous chip generations.
According to Apple, the new Mac mini with M6 delivers up to a 40% increase in CPU performance, up to 4 times faster AI tasks, twice as fast graphics and twice as fast storage compared with the previous M4 model with 32 GB of memory. Memory bandwidth has increased to 170 GB/s versus 153 GB/s in the M5.
In addition to the M6 version, a configuration with the M5 Pro chip is also available — a 15-core CPU and 16-core GPU by default, with the option to upgrade to an 18-core CPU and 20-core GPU.
Mac Studio M5 Ultra: Up to 512 GB of Memory
In parallel, Apple updated the Mac Studio, introducing the M5 Ultra chip — the company’s first quad-die architecture, which combines four M5 Max chips in a single package. This makes it possible to reach a maximum of 512 GB of unified memory — a level previously available only on the M3 Ultra, before Apple quietly removed this option from the configurator in early 2026.
This amount of memory is critically important for developers who run large language models locally without relying on cloud servers.
Why This Matters Right Now
Both devices — Mac mini and Mac Studio — have become especially popular over the past year among users running autonomous AI agents locally, such as OpenClaw. Demand was so high that previous-generation models were being sold with price markups on marketplaces like eBay.
The new Mac mini starts at $899 — $100 more than the previous model. This increase is partly related to the general rise in memory chip prices caused by the global shortage driven by demand from AI data centers — the same factor that already forced Apple to raise prices on a number of other devices this summer.
The new products will begin shipping on September 22 alongside macOS 27.
In Brief: The Key Points
- August 25, 2026 — Mac mini with the M6 chip, Apple’s first 2 nm chip, was released
- Dual Neural Engine — two 16-core blocks working simultaneously, for the first time in Apple’s history
- Up to 4 times faster AI tasks compared with M4
- Mac Studio with M5 Ultra — up to 512 GB of unified memory, quad-die architecture
- Mac mini price — from $899, $100 more than the previous model
- Shipping begins on September 22 alongside macOS 27
FAQ
When was the Mac mini M6 released? On August 25, 2026. Shipping begins on September 22 alongside macOS 27.
How much does the new Mac mini cost? The version with the M6 chip starts at $899, while the version with M5 Pro starts at $1,699.
What makes the M6 chip special? It is Apple’s first chip manufactured using a 2-nanometer process, with a dual Neural Engine — two 16-core blocks working simultaneously instead of one in previous generations.
How much memory does the Mac Studio with M5 Ultra support? Up to 512 GB of unified memory — thanks to the new quad-die architecture that combines four M5 Max chips.
Why did the Mac mini become $100 more expensive? Partly because of the overall rise in memory chip prices amid a global shortage driven by demand from AI data centers.
The article was prepared by the TechVisor team — practical IT media for people.
