M2 Max 96GB vs RTX 5090

For running local LLMs · 4 reports across 2 models

Side A

Vendor: apple
VRAM: 96GB
Memory: Unified

Side B

Vendor: nvidia
VRAM: 32GB
Memory: Discrete

Tokens per second by model

Model	M2 Max 96GB	RTX 5090
Qwen3.6up to 35B	—	106.5n=3
Gemma 4up to 31B	—	—n=1

More comparisons