Apple's $899 M6 Mac Mini Is Openly Built for Running LLMs Locally
Scobleizer · x · 2026-08-26
Apple is no longer hiding it: the new Mac Mini with M6, its first 2nm chip, starts at $899 and is pitched for local AI.
- $899 Mac Mini (32GB): comfortably runs Qwen 3.6 27B (near-GPT-4 quality), Qwen 3 Coder 32B as a full desk-side coding assistant, full-precision Llama 4 8B, and DeepSeek R1 32B; LLM processing is 13.5x faster than M1 and 4x faster than M4.
- Mac Studio with M5 Ultra (512GB): runs Llama 4 70B at high-quality quantization, DeepSeek V4 with hundreds of billions of parameters, and frontier-scale models entirely in local memory; four Mac Studios can be clustered over Thunderbolt 5 to run trillion-parameter models locally.
Apple said Mac grew 30% last quarter largely because people are buying these to run AI models on their own desks instead of paying for cloud.
More from Embodied
- Humanoid Robot TianGong Runs 100m in 8.86 Seconds, Breaking Record — CSProfKGD · 2026-08-26
- Figure pledges over $1B for data and compute to scale Helix robot dataset — coreylynch · 2026-08-26
- Apple Unveils M6 Mac Mini Starting at $899 — soumitrashukla9 · 2026-08-26
- New Robot Competitor SkildAI Enters the Ring, Challenging Figure and Tesla — oouuaaii · 2026-08-26
- Figure's Index launches as world's largest robot dataset, $15M paid out, $1B committed — adcock_brett · 2026-08-26
- Vibe-coding a demo isn't enough for real field robots — ZeYanjie · 2026-08-26