can-i-run-this-llm

Can I run this LLM?

Can a MacBook Air M2 (8GB) run Llama 3.1 8B?

Llama · Llama 3.1 Community · estimated for a 8.0 GB machine at 4K context.

Yes — it runs great.

Best fit: Llama 3.1 8B at Q3_K_M · 3.5 GB · 15–23 tok/s on a MacBook Air M2 (8GB).

Every quantization on this machine

QuantSizeSpeedFit
Q2_K2.6 GB19–29 tok/sRuns great
Q3_K_M3.5 GB15–23 tok/sRuns great
Q4_K_M4.9 GB12–17 tok/sWon't fit
Q5_K_M5.7 GB10–15 tok/sWon't fit
Q6_K6.6 GB9–13 tok/sWon't fit
Q8_08.5 GB7–11 tok/sWon't fit

Ran Llama 3.1 8B on a MacBook Air M2 (8GB)? Add your real numbers →