M5 Max 64GB vs RTX 4070

For running local LLMs · 3 reports across 2 models

Side A

Vendor: apple
VRAM: 64GB
Memory: Unified

Side B

Vendor: nvidia
VRAM: 12GB
Memory: Discrete

Tokens per second by model

Model	M5 Max 64GB	RTX 4070
Gemma 4up to 31B	—	55.0n=2
Qwen3.6up to 35B	32.0n=1	—

More comparisons