GPU

Apple Mac Studio M4 Ultra (80-core GPU) 128GB for local AI

96GB memory, 820 GB/s bandwidth, and deterministic compatibility links for downloadable local models.

VRAM

96GB

Bandwidth

820 GB/s

Architecture

Apple M4 Ultra

Memory type

LPDDR5X Unified

Models to check on this GPU

• Can it run command-a-plus-05-2026-w4a4?
• Can it run Qwen3.5-122B-A10B?
• Can it run NVIDIA-Nemotron-3-Super-120B-A12B-BF16?
• Can it run OpenAI gpt-oss-120b?
• Can it run GLM-4.5-Air?
• Can it run GLM-4.5V?
• Can it run GLM-4.6V?
• Can it run Qwen3-Next-80B-A3B-Thinking?