MacBook Air M2 16GB

Apple · M2 · 16GB Unified Memory · Can run 36 models

Buy Apple
Manufacturer Apple
Unified Mem 16 GB
Chip M2
CPU Cores 8
GPU Cores 10
Neural Engine 16
Bandwidth 100 GB/s
MSRP $1,299
Released Jul 15, 2022

AI Notes

The MacBook Air M2 16GB is a strong portable option for local AI. With 16GB unified memory and 100 GB/s bandwidth, it runs 7B models quickly and handles 13B models with Q4 quantization at moderate speeds. The fanless design keeps it silent during inference.

Compatible Models

Model Parameters Best Quant VRAM Used Fit Est. Speed
Gemma 3 1B 1B Q8_0 2 GB Runs ~50 tok/s
Llama 3.2 1B 1B Q8_0 3 GB Runs ~33 tok/s
DeepSeek R1 1.5B 1.5B Q8_0 3 GB Runs ~33 tok/s
Gemma 2 2B 2B Q8_0 4 GB Runs ~25 tok/s
Llama 3.2 3B 3B Q8_0 5 GB Runs ~20 tok/s
Phi-3 Mini 3.8B 3.8B Q8_0 5.8 GB Runs ~17 tok/s
Phi-4 Mini 3.8B 3.8B Q4_K_M 4.5 GB Runs ~22 tok/s
Gemma 3 4B 4B Q4_K_M 5 GB Runs ~20 tok/s
Qwen 3 4B 4B Q4_K_M 4.5 GB Runs ~22 tok/s
DeepSeek R1 7B 7B Q8_0 9 GB Runs ~11 tok/s
Mistral 7B 7B Q8_0 9 GB Runs ~11 tok/s
Qwen 2.5 7B 7B Q8_0 9 GB Runs ~11 tok/s
Qwen 2.5 Coder 7B 7B Q8_0 9 GB Runs ~11 tok/s
DeepSeek R1 8B 8B Q4_K_M 7.5 GB Runs ~13 tok/s
Llama 3.1 8B 8B Q8_0 10 GB Runs ~10 tok/s
Qwen 3 8B 8B Q4_K_M 7.5 GB Runs ~13 tok/s
Gemma 2 9B 9B Q8_0 11 GB Runs ~9 tok/s
Gemma 3 12B 12B Q4_K_M 10.5 GB Runs ~10 tok/s
Mistral Nemo 12B 12B Q4_K_M 9.5 GB Runs ~11 tok/s
DeepSeek R1 14B 14B Q4_K_M 9.9 GB Runs ~10 tok/s
Phi-4 14B 14B Q4_K_M 9.9 GB Runs ~10 tok/s
Qwen 2.5 14B 14B Q4_K_M 9.9 GB Runs ~10 tok/s
Qwen 2.5 Coder 14B 14B Q4_K_M 12 GB Runs ~8 tok/s
Qwen 3 14B 14B Q4_K_M 12 GB Runs ~8 tok/s
Codestral 22B 22B Q4_K_M 14.7 GB Runs (tight) ~7 tok/s
StarCoder2 15B 15B Q8_0 17 GB CPU Offload ~6 tok/s
Devstral 24B 24B Q4_K_M 17 GB CPU Offload ~6 tok/s
Mistral Small 3.1 24B 24B Q4_K_M 18 GB CPU Offload ~6 tok/s
Gemma 2 27B 27B Q4_K_M 17.7 GB CPU Offload ~6 tok/s
Gemma 3 27B 27B Q4_K_M 20 GB CPU Offload ~5 tok/s
Qwen 3 30B-A3B (MoE) 30B Q4_K_M 22 GB CPU Offload ~5 tok/s
DeepSeek R1 32B 32B Q4_K_M 20.7 GB CPU Offload ~5 tok/s
Qwen 2.5 32B 32B Q4_K_M 20.7 GB CPU Offload ~5 tok/s
Qwen 2.5 Coder 32B 32B Q4_K_M 23 GB CPU Offload ~4 tok/s
Qwen 3 32B 32B Q4_K_M 23 GB CPU Offload ~4 tok/s
Command R 35B 35B Q4_K_M 22.5 GB CPU Offload ~4 tok/s
7 model(s) are too large for this hardware.