What LLMs Can the Apple M4 Max 36GB Run?

The Apple M4 Max 36GB has 36 GB of unified memory (about 24.0 GB usable by the GPU) and 410 GB/s of bandwidth. It runs 41 of 58 models fully at Q4_K_M with 8K context, up to Qwen 3.5 35B-A3B, and generates about 49 tokens/s on an 8B model.

Apple · Apple Silicon · Released 2024-10 · Updated October 14, 2026

41 of 58 models fit

Apple M4 Max 36GB (24.0 GB GPU-usable) at Q4_K_M with 8K context. Largest that fits fully: Qwen 3.5 35B-A3B.

ModelNeedsStatusSpeedBest quant that fits
Qwen 2.5 32B21.4 GBFits~13 tok/sQ4_K_M
Qwen 2.5 Coder 32B21.4 GBFits~13 tok/sQ4_K_M
DeepSeek R1 Distill Qwen 32B21.4 GBFits~13 tok/sQ4_K_M
Qwen 3 32B21.4 GBFits~13 tok/sQ4_K_M
Qwen 3.5 35B-A3B21.0 GBFits~88 tok/sQ4_K_M
Qwen 3.6 35B-A3B21.0 GBFits~88 tok/sQ4_K_M
Gemma 4 31B19.8 GBFits~13 tok/sQ4_K_M
Qwen 3 30B-A3B18.8 GBFits~72 tok/sQ5_K_M
Qwen3-Coder 30B-A3B18.8 GBFits~72 tok/sQ5_K_M
Gemma 3 27B17.6 GBFits~15 tok/sQ5_K_M
Qwen 3.5 27B16.9 GBFits~16 tok/sQ6_K
Qwen 3.6 27B16.9 GBFits~16 tok/sQ6_K
Qwen 3.8 27B16.9 GBFits~16 tok/sQ6_K
Mistral Small 3.1 24B15.6 GBFits~17 tok/sQ6_K
Gemma 4 26B-A4B15.6 GBFits~78 tok/sQ6_K
gpt-oss-20b12.9 GBFits~91 tok/sMXFP4 (official)
Qwen 2.5 14B10.7 GBFits~27 tok/sQ8_0
DeepSeek R1 Distill Qwen 14B10.7 GBFits~27 tok/sQ8_0
Phi-4 14B10.6 GBFits~27 tok/sQ8_0
Qwen 3 14B10.5 GBFits~27 tok/sQ8_0
Ministral 3 14B9.7 GBFits~30 tok/sQ8_0
Mistral Nemo 12B9.0 GBFits~33 tok/sQ8_0
Gemma 3 12B8.8 GBFits~33 tok/sQ8_0
Ministral 3 8B6.7 GBFits~46 tok/sFP16 / BF16
Qwen 3 8B6.6 GBFits~47 tok/sFP16 / BF16
Qwen 3.5 9B6.4 GBFits~47 tok/sFP16 / BF16
Llama 3.1 8B6.3 GBFits~49 tok/sFP16 / BF16
DeepSeek R1 Distill Llama 8B6.3 GBFits~49 tok/sFP16 / BF16
Mistral 7B v0.35.7 GBFits~53 tok/sFP16 / BF16
Gemma 4 E4B5.7 GBFits~53 tok/sFP16 / BF16
Qwen 2.5 7B5.6 GBFits~54 tok/sFP16 / BF16
DeepSeek R1 Distill Qwen 7B5.6 GBFits~54 tok/sFP16 / BF16
Qwen 3 4B4.2 GBFits~85 tok/sFP16 / BF16
Phi-4 Mini4.1 GBFits~90 tok/sFP16 / BF16
Gemma 4 E2B4.0 GBFits~83 tok/sFP16 / BF16
Gemma 3 4B3.7 GBFits~92 tok/sFP16 / BF16
Qwen 3.5 4B3.7 GBFits~95 tok/sFP16 / BF16
Ministral 3 3B3.5 GBFits~103 tok/sFP16 / BF16
Llama 3.2 3B3.5 GBFits~107 tok/sFP16 / BF16
Llama 3.2 1B1.7 GBFits~287 tok/sFP16 / BF16
Gemma 3 1B1.6 GBFits~379 tok/sFP16 / BF16
Kimi K2.6566 GBToo small––
DeepSeek R1 671B380 GBToo small––
Llama 3.1 405B234 GBToo small––
Qwen 3.5 397B-A17B226 GBToo small––
DeepSeek V4 Flash161 GBToo small––
Qwen 3 235B-A22B135 GBToo small––
MiniMax-M2133 GBToo small––
Qwen 3.5 122B-A10B70.2 GBToo small––
Mistral Small 4 119B68.2 GBToo small––
gpt-oss-120b67.2 GBToo small––
Llama 4 Scout64.0 GBToo small––
GLM-4.5-Air62.1 GBToo small––
Qwen3-Next 80B-A3B46.3 GBToo small––
Qwen 2.5 72B44.5 GBToo small––
Llama 3.3 70B43.2 GBToo small––
DeepSeek R1 Distill Llama 70B43.2 GBToo small––
Mixtral 8x7B28.0 GBToo small–Q2_K

Apple M4 Max 36GB for local AI

The entry M4 Max (14-core CPU) with 410 GB/s. About 24 GB is GPU-usable by default, similar to a 24 GB GPU for model size but slower per token. Raising the wired-memory limit gives 32B models room for longer contexts.

Among the 8 devices we track with about 24 GB of usable memory, the Apple M4 Max 36GB ranks 6 of 8 for text-generation speed, because generation speed follows memory bandwidth. Others in the same class: RTX 3090 Ti 24GB, RTX 4090 24GB, Radeon RX 7900 XTX 24GB, RTX 3090 24GB. They all hold the same models; the faster ones simply answer sooner.

Our picks for the Apple M4 Max 36GB

The model we would run first for each job, chosen from our recommendations by what fits fully with an 8K context. "Best quant" is the highest quality that fits; "max context" is at Q4_K_M or the native format.

TaskModelBest quantMax contextSpeed
Best all-rounderQwen 3.6 27BQ6_K98K~16 tok/s
CodingQwen 3.6 27BQ6_K98K~16 tok/s
Fastest repliesQwen 3.6 35B-A3BQ4_K_M88K~88 tok/s
Long documents (64K+)Qwen 3.8 27BQ6_K98K~16 tok/s
Images in, text outGemma 4 31BQ4_K_M43K~13 tok/s
Image generationQwen-ImageFP8––
SpecValue
Memory36 GB unified (24.0 GB GPU-usable by default)
Memory bandwidth410 GB/s
ArchitectureApple Silicon
Released2024-10
Llama 3.1 8B Q4_K_M speed~49 tok/s
Largest model that fitsQwen 3.5 35B-A3B

Image and video models on the Apple M4 Max 36GB

Peak VRAM with model swapping (ComfyUI default on smaller cards).

ModelFP8FP8 fitGGUF Q4Q4 fit
Stable Diffusion 1.51.8 GBFits1.7 GBFits
SDXL 1.04.1 GBFits4.1 GBFits
Stable Diffusion 3.5 Medium6.4 GBFits6.4 GBFits
Stable Diffusion 3.5 Large8.6 GBFits6.4 GBFits
FLUX.1 [dev]12.1 GBFits7.3 GBFits
FLUX.1 [schnell]12.1 GBFits7.3 GBFits
Qwen-Image20.4 GBFits12.3 GBFits
Wan 2.2 A14B (T2V / I2V)18.8 GBFits13.2 GBFits
Wan 2.2 TI2V 5B8.0 GBFits6.5 GBFits
HunyuanVideo18.3 GBFits13.1 GBFits
FLUX.1 Kontext [dev]12.3 GBFits7.5 GBFits
FLUX.2 [dev]31.0 GBToo small23.6 GBTight fit
Z-Image Turbo6.8 GBFits4.9 GBFits
Qwen-Image-Edit 250920.5 GBFits12.4 GBFits
LTX-219.2 GBFits12.6 GBFits

Upgrade path

The next step up is the Apple M4 Pro 48GB with 48 GB.

Apple M4 Max 36GB for LLMs: common questions

Can the Apple M4 Max 36GB run Llama 3.3 70B?

No. It needs about 43.2 GB at Q4_K_M, far more than the 24.0 GB available.

Can the Apple M4 Max 36GB run Qwen 3 32B?

Yes. Qwen 3 32B needs about 21.4 GB at Q4_K_M with 8K context, and the best quality that fits is Q4_K_M.

Can the Apple M4 Max 36GB run Qwen 3 14B?

Yes. Qwen 3 14B needs about 10.5 GB at Q4_K_M with 8K context, and the best quality that fits is Q8_0.

How fast is the Apple M4 Max 36GB for LLMs?

About 49 tokens per second for an 8B model at Q4_K_M, based on its 410 GB/s memory bandwidth. Larger models scale down roughly in proportion to their size.

How much of the 36 GB can the GPU use?

About 24.0 GB by default. You can raise it with sudo sysctl iogpu.wired_limit_mb=<MB> (resets on reboot), leaving at least 8 GB for macOS.

Related