LocalAI

See the speed before choosing hardware

Pre-release · 32GBScheduled for release on September 22, 2026Updated September 2026

Mac mini M6 32GB If you want 32GB for a 27B model, check the memory headroom before the new chip.

The 32GB option is worth comparing if you plan to run a Qwen-class 27B Q4 model alongside a browser or document app. It has the same 170GB/s memory bandwidth as the M6 24GB, so the reason to buy the extra 8GB should be room left after loading the model. Launch is planned for September 22; the generation times shown here are estimates, not measurements of a retail unit.

Check Apple's official release schedule

Hardware core specifications

available memory
27GB
Memory bandwidth
170GB/s
Estimated sustained LLM load
About 55W
Korean price range (KRW)
Check the 32GB configuration price

a person who fits well

  • Run a 27B Q4 model alongside work apps
  • Start with short prompts and extend context
  • A compact Mac without building a PC

In this case skip it

If you only plan to use a 12B-class model with short prompts, there is no need to order 32GB first. Check whether the model and prompt cache fit in 24GB. If CUDA-only apps or nodes are essential, more unified memory will not make them compatible. Recheck retail configurations and measured results after the September 22 launch.

Recommended format for each device 4K input

The model you will experience with this equipment

MiniCPM5-2B

Required 2.07GB · spare time

MLX 4-bit

Decode

79.9 ~ 89 tok/s

First token

2.4 to 6.0 seconds

speed tuning

Qwen3.8 27B

recommended

Required 17.56GB · spare time

MLX 4-bit

Decode

7.5 ~ 8.3 tok/s

First token

11.4 ~ 25.0 s

speed tuning

Qwen3.6 35B-A3B (MoE)

Required 21.28GB · titration

MLX 4-bit

Decode

11 ~ 28.7 tok/s

First token

4.3 ~ 11.2 s

speed tuning

Actual speed may vary depending on runtime version, cooling status, context length and quantization file.

purchase judgment

What tasks is it suitable for?

27B conversation

Conditional

Start with Q4 and a 4K input. Even if the model fits, long conversations or other open apps reduce room for the prompt cache. Pre-release figures are estimates.

Documents and coding

Suitable

Compare the length of the sections you actually need instead of sending an entire document by default. Long inputs affect the wait for the first token; long answers affect generation time.

Image and CUDA workflows

Conditional

You can use models and nodes supported on Mac, but CUDA-only extensions will not transfer unchanged. Check your existing workflow first.

The first thing to get stuck
The model cannot use all 32GB: macOS, open apps, and prompt cache share the unified memory. Apple's specifications give the M6 24GB and 32GB the same 170GB/s bandwidth, so when a workload fits on both, the capacity difference alone will not increase generation speed proportionally.
Configuration to check
M6 · 12-core GPU · 32GB unified memory · 512GB or larger SSD recommended (check retail options)
Criteria for spending more money
If you already own an M4 32GB, memory capacity alone is not a reason to replace it with an M6. Compare prompt processing and long-answer generation using the same model, quantization, and input length. For a new purchase, first check whether the M6 24GB can do the job; pay for 32GB only if a 27B model plus your other apps leave too little room. If waiting time is the main problem, the M4 Pro 48GB is also worth comparing.

Check the official M6 32GB configuration and launch date.

This product is available for pre-order. Apple says sales begin September 22. Confirm the exact 32GB and SSD options and the delivery date.

Check the official release schedule

Frequently Asked Questions

When will the Mac mini M6 32GB be delivered?
Apple says the new Mac mini will begin arriving to customers and appearing in stores on September 22, 2026. Check Apple's order page for the 32GB option's price and delivery date.
What LLM can be run on the Mac mini M6 32GB?
You can compare MiniCPM5-2B, Qwen3.8 27B, Qwen3.6 35B-A3B (MoE), etc. with 4K input and recommended formats. Memory requirements vary depending on model and context length.
What is the local LLM power of the Mac mini M6 32GB?
Expected to be around 55W under sustained load, with an actual range estimated between 35–80W.
Are the displayed token speeds ground truth?
This is an estimated range calibrated to the calculator against official hardware specifications and verified benchmark ranges. Runtime, cooling, quantization may vary depending on file and context length.

Equipment to compare together

Consider less memory for mainly 12B models

Mac mini M6 24GB

M6 24GB configuration with 170GB/s memory bandwidth. You can compare the 12B class with some 27B Q4 models.

Compare with the previous 32GB generation

Mac mini M4 32GB

Consider this configuration if you want to run 27B Q4 without building a new PC. Moving up makes sense if 24GB was running out of memory. If the model and cache already fit, however, choosing 32GB does not make the chip itself faster.

If waiting time matters more

Mac mini M4 Pro 48GB

When you repeatedly read documents and edit code with a 27B model, each wait matters as well as capacity. This page uses the 20-core GPU, 48GB configuration. Distinguish it from less expensive listings with a 16-core GPU.