← Back to events
ActiveTech

Run Qwen3.8 27B locally: real numbers from my Mac Studio

Photo: Hacker News

What happened

Qwen3.8 27B benchmarked on a Mac Studio M3 Ultra: Ollama and llama.cpp numbers, the 1-bit quant tested, RAM math, and the hardware that actually fits.

Summary assembled by rule from the sources below

Why it's spreading

Sources