Can I Run GLM-4.7 / GLM-Z1 on Mac Studio (M3 Ultra, 256 GB)?

Written by Jakub Rusinowski · Last updated August 15, 2026

Mac Studio (M3 Ultra, 256 GB) — what it gives a model

Usable memory for models	192 GB
Memory bandwidth	819 GB/s
Form factor	Desktop
Operating system	macOS
Memory upgradeable	No — soldered

GLM-4.7 / GLM-Z1 on Mac Studio (M3 Ultra, 256 GB): memory by quantization

Quant	Memory needed	Fits 192 GB?	Max context	Est. speed	Download
F16	66.6 GB	✓ Yes	64K	~6.8 tok/s	64 GB
Q8_0	36.6 GB	✓ Yes	64K	~12.4 tok/s	34 GB
Q6_K	28.9 GB	✓ Yes	64K	~15.7 tok/s	26.2 GB
Q5_K_M	25.3 GB	✓ Yes	64K	~18 tok/s	22.7 GB
Q4_K_M	22 GB	✓ Yes	64K	~20.8 tok/s	19.3 GB
Q3_K_M	16.3 GB	✓ Yes	64K	~28.2 tok/s	13.6 GB
Q2_K	13.2 GB	✓ Yes	64K	~35.1 tok/s	10.5 GB

Which GLM-4.7 / GLM-Z1 sizes fit

Variant	Needs at Q4_K_M	Fits?	Est. speed
GLM-Z1 32B (Reasoning)	22 GB	✓ Fits	~20.8 tok/s
GLM-4.7-Flash 30B-A3B	20.7 GB	✓ Fits	~110.9 tok/s
GLM-4.7 9B	7.4 GB	✓ Fits	~60.8 tok/s

What to watch out for

• This model does not publish its full attention configuration, so the KV-cache share of these figures is inferred from its parameter count rather than computed exactly.
• Memory on this machine is not upgradeable, so the configuration you buy is the ceiling for every model you will ever run on it.

Mac Studio M3 Ultra 256 GB limitations

• 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.

Recommended setup

Ollama or LM Studio (Metal) — MLX for the fastest Apple-native throughput

How these numbers are calculated

• Memory need = quantized weights + KV cache at 8,192 tokens (f16) + 0.8 GB runtime overhead.
• 256 GB unified memory at 819 GB/s, shared between CPU and GPU.
• macOS reserves a share of unified memory for the system, so not all of it is available to a model.
• Throughput is a memory-bandwidth roofline estimate, not a measurement. It is labelled "estimated" everywhere it appears.
• KV cache is inferred from the parameter count because this model does not publish its full attention configuration.

FAQ

Can I run GLM-4.7 / GLM-Z1 on Mac Studio (M3 Ultra, 256 GB)?

Yes, comfortably — GLM-Z1 32B (Reasoning) at Q8_0 needs about 36.6 GB of the 192 GB usable on Mac Studio (M3 Ultra, 256 GB), leaving ~155.4 GB spare and running at ~12.4 tok/s (estimated), with room for about 65,536 tokens of context.

Which quantization of GLM-4.7 / GLM-Z1 should I use on Mac Studio (M3 Ultra, 256 GB)?

Q8_0 — it needs about 36.6 GB of the 192 GB available, downloads as roughly 34 GB, and runs at an estimated 12.4 tokens/sec with up to 64K of context.

What limits GLM-4.7 / GLM-Z1 on Mac Studio (M3 Ultra, 256 GB)?

Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.

Which runtime should I use?

Ollama or LM Studio (Metal) — MLX for the fastest Apple-native throughput

Other Computers

• GLM-4.7 / GLM-Z1 on RTX 4090 desktop
• GLM-4.7 / GLM-Z1 on RTX 5090 desktop
• GLM-4.7 / GLM-Z1 on RTX 3090 desktop
• GLM-4.7 / GLM-Z1 on Mac Studio M4 Max 64 GB

Other Models on Mac Studio (M3 Ultra, 256 GB)

• GLM-5 / GLM-5.1 on Mac Studio (M3 Ultra, 256 GB)
• GLM-6 on Mac Studio (M3 Ultra, 256 GB)
• GPT-OSS on Mac Studio (M3 Ultra, 256 GB)
• Granite 3.0 on Mac Studio (M3 Ultra, 256 GB)
• IBM Granite 4.0 on Mac Studio (M3 Ultra, 256 GB)

GLM-4.7 / GLM-Z1 on GPUs

• GLM-4.7 / GLM-Z1 on NVIDIA GeForce RTX 5090
• GLM-4.7 / GLM-Z1 on NVIDIA GeForce RTX 5070
• GLM-4.7 / GLM-Z1 on NVIDIA GeForce RTX 5060 Ti 8GB
• GLM-4.7 / GLM-Z1 on NVIDIA GeForce RTX 5060

What This Model Is Good At

• Best local LLMs for reasoning
• Best local LLMs for research
• Best local LLMs for mathematics

Model & Tools

• GLM-4.7 / GLM-Z1 model page
• VRAM calculator
• Check your own hardware

← Can I Run It? | GLM-4.7 / GLM-Z1 model page | Check your hardware