M5 Max (18-core CPU / 32-core GPU)

Name: M5 Max (18-core CPU / 32-core GPU)
Brand: Apple
Price: 84900 TWD

Apple M5 Max (18-core CPU / 32-core GPU): 36 GB of memory at 460 GB/s. Below: the open models it can run, estimated at 36 GB with an 8K context. The binned M5 Max: 460 GB/s but fixed at 36 GB, enough for 32B models at 5-bit.

Last updated: 2026-10-14

Specifications

Memory	36 GB LPDDR5X[src]
Memory bandwidth	460 GB/s[src]
Launch	2026-03[src]
Launch price	US$3,599[src]
Architecture	Apple M5 Max (Fusion Architecture: two third-generation 3 nm dies)[src]
Software	Metal / MLX

What it can run

Best-quality quantization that fits with a 8K context, and the estimated generation speed. Estimates, not benchmarks. See the methodology.

Model	Params	Best quant	Memory	tok/s (est.)
Qwen3.5-35B-A3B	35B (3B)	Q5_K_M	25.0 GiB	140
Qwen3.6-35B-A3B	35B (3B)	Q5_K_M	25.0 GiB	140
Qwen3-32B	32.8B	Q5_K_M	25.3 GiB	13
DeepSeek-R1-Distill-Qwen-32B	32.76B	Q5_K_M	25.3 GiB	13
Gemma 4 31B IT	30.7B	Q6_K	26.3 GiB	12
Qwen3-30B-A3B (2507)	30.5B (3.3B)	Q6_K	25.8 GiB	110
Qwen3-Coder-30B-A3B-Instruct	30.5B (3.3B)	Q6_K	25.8 GiB	110
GLM-4.7-Flash (30B-A3B)	30B (3B)	Q6_K	25.1 GiB	121
Qwen3.5-27B	27B	Q6_K	22.7 GiB	13
Qwen3.6-27B	27B	Q6_K	22.7 GiB	13
Qwen3.8-27B	27B	Q6_K	22.7 GiB	13
Gemma 3 27B IT	27B	Q6_K	23.3 GiB	13
Gemma 4 26B A4B IT	25.2B (3.8B)	Q8_0	27.0 GiB	74
Mistral Small 3.2 24B Instruct (2506)	24B	Q8_0	26.7 GiB	12
Magistral Small 1.2 (2509)	24B	Q8_0	26.7 GiB	12
Devstral Small 2 24B Instruct (2512)	24B	Q8_0	26.7 GiB	12
gpt-oss-20b	21B (3.6B)	Q8_0	22.5 GiB	78
Phi-4 (14B)	14B	Q8_0	16.6 GiB	20
Gemma 3 12B IT	12B	FP16	24.8 GiB	12
Gemma 4 12B Unified IT	11.95B	FP16	24.2 GiB	13
Qwen3.5-9B	9B	FP16	18.4 GiB	17
Qwen3-8B	8.2B	FP16	17.7 GiB	18
DeepSeek-R1-0528-Qwen3-8B	8.19B	FP16	17.6 GiB	18
Llama 3.1 8B Instruct	8B	FP16	17.1 GiB	19
Gemma 4 E4B IT	8B	FP16	16.4 GiB	19
Llama 3.2 3B Instruct	3.21B	FP16	7.7 GiB	47
Kimi K2.6	1000B (32B)	Doesn't fit	–	–
DeepSeek-R1-0528 (671B)	671B (37B)	Doesn't fit	–	–
DeepSeek-V4-Flash	284B (13B)	Doesn't fit	–	–
Qwen3.8-Flash-Next	180B (6B)	Doesn't fit	–	–
Qwen3.5-122B-A10B	122B (10B)	Doesn't fit	–	–
Mistral Small 4 119B (2603)	119B (6.5B)	Doesn't fit	–	–
gpt-oss-120b	117B (5.1B)	Doesn't fit	–	–
Llama 4 Scout 17B-16E Instruct	109B (17B)	Doesn't fit	–	–
GLM-4.5-Air	106B (12B)	Doesn't fit	–	–
Qwen3-Coder-Next (80B-A3B)	80B (3B)	Doesn't fit	–	–
Llama 3.3 70B Instruct	70B	Doesn't fit	–	–

Prices

Latest price at each shop, as listed on the day we checked.

Region	Shop	Listing	Price	Per GB	Checked
Taiwan	Apple	Mac Studio，M5 Max 晶片，18 核心 CPU、32 核心 GPU，36GB 統一記憶體，512GB 儲存裝置Price shown in Apple Store configurator for this configuration	NT$84,900	NT$2,358/GB	2026-10-14
US	Apple	Mac Studio, M5 Max chip, 18-core CPU, 32-core GPU, 36GB memory, 512GB storage	US$2,499	US$69/GB	2026-10-14

Taiwan value: about 61.6 tok/s on Llama 3.1 8B Instruct (Q4_K_M, 8K), or 0.7 tok/s per NT$1,000.
US value: about 61.6 tok/s on Llama 3.1 8B Instruct (Q4_K_M, 8K), or 2.5 tok/s per US$100.

A price history chart appears once we have checked this device on two or more days.

All prices →

Frequently asked questions

Which LLMs can the M5 Max (18-core CPU / 32-core GPU) run?

With 36 GB and an 8K context, the largest models that fit include: Qwen3.5-35B-A3B, Qwen3.6-35B-A3B, Qwen3-32B, DeepSeek-R1-Distill-Qwen-32B, Gemma 4 31B IT. See the full table below.

How fast is the M5 Max (18-core CPU / 32-core GPU) for local LLMs?

Generation speed is mostly set by memory bandwidth (460 GB/s). For Llama 3.1 8B Instruct at FP16, expect roughly 19 tokens/s; bigger models are slower.

Can the M5 Max (18-core CPU / 32-core GPU) run 70B models?

Llama 3.3 70B Instruct does not fit on the M5 Max (18-core CPU / 32-core GPU) (36 GB), even at Q4_K_M with an 8K context.