M3 8GB vs M5 Max 128GB

For running local LLMs · 3 reports across 2 models

Side A

Vendor: apple
VRAM: 8GB
Memory: Unified

Side B

Vendor: apple
VRAM: 128GB
Memory: Unified

Tokens per second by model

Model	M3 8GB	M5 Max 128GB
Gemma 4up to 31B	18.0n=1	7.5n=1
Qwen3.6up to 35B	—	5.5n=1

More comparisons