Mac Studio (M3 Ultra, 256 GB) — Local LLM Performance

Mac Studio (M3 Ultra, 256 GB) is a Desktop running macOS, with 256 GB unified memory.

Specifications

Memory for models	256 GB unified
Memory bandwidth	819 GB/s
Memory upgradeable	No — soldered, decided at purchase
Chip	Apple M3 Ultra
Architecture	ARM, 3nm TSMC
Form factor	Desktop
Operating system	macOS
Released	2025
Llama 3.1 8B Q4_K_M	43–90 tok/s

Worth knowing before you buy

• The largest single-box unified memory pool available, but at roughly half the memory bandwidth of a high-end discrete GPU — very large models load, then run slowly.
• Memory is soldered; the configuration chosen at purchase is permanent.

Models on the Mac Studio M3 Ultra 256 GB

• Yi 1.5 Family — Yes — comfortably
• VibeThinker — Yes — comfortably
• Aya 3B (Tiny Aya) — Yes — comfortably
• StarCoder 2 — Yes — comfortably
• SmolLM3 — Yes — comfortably
• SmolLM2 — Yes — comfortably
• Qwen3-Coder — Yes — comfortably
• Qwen3.8 — Yes — comfortably
• Qwen 3.7 — Yes — comfortably
• Qwen 3.6 — Yes — comfortably
• Qwen 3.5 — Yes — comfortably
• Qwen 3 — Yes
• Qwen 2.5 VL — Yes — comfortably
• Qwen 2.5 Family — Yes — comfortably
• Poolside Laguna XS 2.1 — Yes — comfortably
• Phi-4 Mini — Yes — comfortably
• Phi-4 Family — Yes — comfortably
• Phi 3.5 Family — Yes — comfortably
• OLMo 2 — Yes — comfortably
• Nemotron Cascade 2 — Yes — comfortably
• Nemotron 70B — Yes — comfortably
• Nemotron 3 Super — Yes — comfortably
• Nemotron 3 Nano Omni — Yes — comfortably
• Mistral Small 4 — Yes — comfortably
• Mistral Small 3.1 — Yes — comfortably
• Mistral Small 3.2 — Yes — comfortably
• Mistral Family — Yes — comfortably
• Ministral — Yes — comfortably
• Ministral 3 — Yes — comfortably
• MiniMax M3 — Yes
• MiniMax M2.7 — Yes
• MiniMax M2.5 — Yes
• MiMo-V2.5 — Yes, but it is tight
• Magistral Small — Yes — comfortably
• Llama 4 — Yes — comfortably
• Llama 4.5 — Yes — comfortably
• Llama 3.3 — Yes — comfortably
• Llama 3.2 Vision — Yes — comfortably
• Llama 3.2 Family — Yes — comfortably
• Llama 3.1 Family — Yes — comfortably
• InternLM 3 — Yes — comfortably
• IBM Granite 4.0 — Yes — comfortably
• IBM Granite 4.1 — Yes — comfortably
• Granite 3.0 — Yes — comfortably
• GPT-OSS — Yes — comfortably
• GLM-4.7 / GLM-Z1 — Yes — comfortably
• GLM-6 — Yes — comfortably
• GLM-5 / GLM-5.1 — Yes — comfortably
• Gemma 4 — Yes — comfortably
• Gemma 3n — Yes — comfortably
• Gemma 3 — Yes — comfortably
• Gemma 2 Family — Yes — comfortably
• Falcon 3 — Yes — comfortably
• EXAONE 3.5 — Yes — comfortably
• Devstral — Yes — comfortably
• DeepSeek V4 — Yes
• DeepSeek V4.1 — Yes
• DeepSeek R1 — Yes — comfortably
• Cosmos 3 — Yes — comfortably
• Command R Family — Yes — comfortably
• Cogito v1 — Yes — comfortably
• Codestral — Yes — comfortably
• Bonsai 27B — Yes — comfortably
• BitNet b1.58 — Yes — comfortably
• Aya Expanse — Yes — comfortably

← All hardware | AI stations and mini-PCs | Check your hardware