The verdict is the new Mac mini M6 (16 GB) at $899 (Apple list, 14 Nov 2026), unless a used M4 (16 GB) costs under about $700. The LLMCheck index puts Gemma 4 12B at ~17 tok/s (estimated) on the new Mac against ~13 tok/s (estimated) on the used one, and both fit the same 38 of the 130 models.
| M4 (16 GB) | M6 (16 GB) | |
|---|---|---|
| Chip | M4 | M6 |
| GPU cores | 10-core | 12-core |
| Memory bandwidth | 120 GB/s | 153 GB/s |
| Unified memory | 16 GB | 16 GB |
| Price | — | $899 |
| Price per GB | — | $56 |
| Models that fit (of 130) | 38 | 38 |
| Best local LLM here | MiniCPM5-2B estimated66 tok/s | Gemma 4 12B estimated17 tok/s |
| Gemma 4 12B | estimated13 tok/s | estimated17 tok/s |
| Price per tok/s on Gemma 4 12B | — | $53 |
| Where to buy | Used / refurbished (Apple no longer sells it new) | New · Amazon stocks this configuration |
Every speed below is the LLMCheck index’s figure for that exact configuration — the same number its Mac mini M4 (16 GB) and Mac mini M6 (16 GB) pages print, with its provenance mark. Across the 34 shared models with a figure on both, the M6 column is a median ×1.25 of the M4 column.
| Model | Size | M4 tok/s | M6 tok/s | Ratio |
|---|---|---|---|---|
| MiniCPM5-2B | 2.52B | estimated66 | estimated82 | ×1.24 |
| Ternary Bonsai 2 27B | 27.36B | — | — | — |
| Maple Preview 20B-A1B caveat | 20B MoE | sourced169 | — | — |
| Bonsai 27B | 27B | — | — | — |
| Gemma 4 12B | 12B | estimated13 slow | estimated17 | ×1.31 |
| Ling-3.0-tiny | 7.9B MoE | estimated115 | estimated127 | ×1.10 |
| LFM3-1.7B | 1.73B | community93 | estimated84 | ×0.90 |
| Nanbeige4.2-3B | 3B | estimated51 | estimated64 | ×1.25 |
| GPT-oss 20B | 21B MoE | estimated34 | estimated41 | ×1.21 |
| Mistral Small 3.2 24B | 24B | estimated7 slow | estimated8 slow | ×1.14 |
| K2-Horizon-3.7B | 5.1B | estimated31 | estimated39 | ×1.26 |
| Devstral Small 24B | 24B | estimated7 slow | estimated8 slow | ×1.14 |
| Spark-X2.6-4B | 4.11B | estimated32 | estimated40 | ×1.25 |
| Mellum2.1-12B-A2.5B-Thinking | 12.15B MoE | estimated46 | estimated55 | ×1.20 |
| MiMo-V2.6-Distill-Qwen-9B | 9.41B | estimated17 | estimated22 | ×1.29 |
Nothing: both fit the same 38 of the 130 models in the LLMCheck index, because both have 16 GB. The difference is speed and price.
The new M6 (16 GB) costs $899 (Apple list, 14 Nov 2026): $56 per GB and $53 per tok/s on Gemma 4 12B. A used M4 (16 GB) matches that $53 per tok/s at about $700; below that price it is the better value on decode speed alone. The index records no used prices: check current listings, and weigh the warranty and the remaining years of macOS support.
Bottom line: The new M6 (16 GB), unless the used one costs under about $700.
Renewed stock of the 16 GB Mac mini M4 varies: check the memory in the listing before you buy.
As an Amazon Associate, LLMCheck earns from qualifying purchases. Affiliate links cost you nothing extra and never influence the verdicts.
Each speed is decode (generation) tok/s from the LLMCheck index for that exact configuration: the lower-median measured run on that chip where one exists (a filled square is a published figure, a hollow circle a community run), otherwise an estimate from the memory-bandwidth model (hollow diamond, printed with a tilde). LLMCheck runs no tests of its own. Prompt processing (prefill) depends on GPU compute rather than bandwidth, and the index does not estimate it; Apple’s M5-generation GPUs add per-core Neural Accelerators aimed at that work.
The verdict is the new Mac mini M6 (16 GB) at $899 (Apple list, 14 Nov 2026), unless a used M4 (16 GB) costs under about $700. The LLMCheck index puts Gemma 4 12B at ~17 tok/s (estimated) on the new Mac against ~13 tok/s (estimated) on the used one, and both fit the same 38 of the 130 models.
Decode speed on Apple Silicon follows memory bandwidth: the Mac mini M6 (16 GB) reads memory at 153 GB/s, the Mac mini M4 (16 GB) at 120 GB/s. The LLMCheck index puts Gemma 4 12B at ~17 tok/s (estimated) against ~13 tok/s (estimated). Across the 34 shared models with a figure on both, the median gain is 1.25x.
Yes. Both fit the same 38 of the 130 models in the LLMCheck index.
About $700 or less. At that price it matches the new Mac's $53 per tok/s on Gemma 4 12B ($899 at Apple list, checked 14 Nov 2026). The LLMCheck index records no used prices, so check current listings, and weigh warranty and macOS support.
On the Mac mini M4 (16 GB): MiniCPM5-2B at ~66 tok/s (estimated). On the Mac mini M6 (16 GB): Gemma 4 12B at ~17 tok/s (estimated). These are the headline picks of the LLMCheck Best by Mac pages: the best-ranked model that reaches 15 tok/s on that Mac and has a standard Mac runtime.