What is the best LLM for an M1 Max with 32 GB?

Of the models with a published Intelligence Index, Qwen3.8 27B scores highest among those that run well here: 4-bit, about 12–21 tokens a second. The biggest that runs well is Qwen3.6 35B A3B.

Can an M1 Max with 32 GB run Llama 3.3 70B?

Yes, but slowly: 2-bit (lower quality), at about 8–15 tokens a second.

How fast is gpt-oss-20b on an M1 Max?

About 95–168 tokens a second (native 4.25-bit), from the chip’s 400 GB/s memory bandwidth.

What LLMs can an M1 Max with 32 GB run? Models and speed

Runs well 38

Fits with room for context, at 10 or more tokens a second.

Model	Version	Memory	Tokens/s
Qwen3.8 27B27B · Intelligence 33.7	4-bit	15.2 GB	12–21
Qwen3.6 35B A3B35B · 3B active · Intelligence 18.2	4-bit	19.7 GB	106–188
Muse Glimmer 30B29.6B · Intelligence 17.5	4-bit	16.7 GB	11–19
Gemma 4 31B30.7B · Intelligence 14.7	4-bit	17.3 GB	10–18
Nemotron 3.5 Lightning 30B A3B30B · 3B active · Intelligence 12.9	4-bit	16.9 GB	106–188
Granite 4.2 30B30B · Intelligence 12.7	4-bit	16.9 GB	11–19
MiniCPM5 2B2.52B · Intelligence 12.5	8-bit	2.7 GB	67–119
Qwen3.5 9B9B · Intelligence 11.2	8-bit	9.6 GB	19–33
Granite 4.2 8B8B · Intelligence 11.1	8-bit	8.5 GB	21–38
Granite 4.2 3B3B · Intelligence 9.1	8-bit	3.2 GB	56–100
gpt-oss-20b21B · 3.6B active · Intelligence 9	native 4.25-bit	13.8 GB	95–168
Nemotron 3 Nano 30B A3B30B · 3.5B active · Intelligence 8.9	4-bit	16.9 GB	90–160
Gemma 4 E4B8B · Intelligence 8.9	8-bit	8.5 GB	38–67
Ministral 3 8B8.9B · Intelligence 5.5	8-bit	9.5 GB	19–34
Nemotron 3 Nano 4B3.97B · Intelligence 4.8	8-bit	4.2 GB	43–76
Qwen3 Coder 30B A3B Instruct30.5B · 3.3B active	4-bit	17.2 GB	95–168
GLM-4.7 Flash30B · 3B active	4-bit	16.9 GB	106–188
Qwen3.6 27B27B	4-bit	15.2 GB	12–21
Gemma 4 26B A4B25.2B · 3.8B active	6-bit	20.5 GB	58–103
Devstral Small 2 (24B)24B	4-bit	13.5 GB	13–24
Magistral Small 1.2 (2509)24B	4-bit	13.5 GB	13–24
Mistral Small 3.2 (24B)24B	4-bit	13.5 GB	13–24
LFM2.5 24B A2B24B · 2B active	6-bit	19.5 GB	113–200
Phi-4 Reasoning Vision 15B15B	8-bit	15.9 GB	11–20
Phi-4 (14B)14B	8-bit	14.9 GB	12–21
Gemma 4 12B Unified12B	8-bit	12.7 GB	14–25
Qwen3 8B8.2B	8-bit	8.7 GB	21–37
Llama 3.1 8B Instruct8B	8-bit	8.5 GB	21–38
Gemma 4 E2B5.1B	8-bit	5.4 GB	75–133
Qwen3.5 4B4B	8-bit	4.3 GB	42–74
Phi-4 mini instruct3.8B	8-bit	4 GB	45–80
Llama 3.2 3B Instruct3.21B	8-bit	3.4 GB	53–94
SmolLM3 3B3B	8-bit	3.2 GB	56–100
LFM2.5 2.6B2.69B	8-bit	2.9 GB	62–110
Qwen3.5 2B2B	8-bit	2.1 GB	86–152
Llama 3.2 1B Instruct1.23B	8-bit	1.3 GB	138–246
Qwen3.5 0.8B0.8B	8-bit	0.9 GB	200–356
Qwen3 0.6B0.6B	8-bit	0.6 GB	300–533

Runs, but slowly 1

Fits, but under 10 tokens a second: fine for batch work, slow for chat.

Model	Version	Memory	Tokens/s
Llama 3.3 70B Instruct70B	2-bit lower quality	21.9 GB	8–15

Only just fits 2

Needs almost all memory: raise the GPU memory limit and keep the context short.

Model	Version	Memory	Tokens/s
Nemotron 3.5 Bolt 74B A9B74B · 9B active · Intelligence 21.4	2-bit lower quality	23.1 GB	64–114
Qwen3 Coder Next80B · 3B active · Intelligence 9.2	2-bit lower quality	25 GB	200–356

Other Macs

Every Mac with 32 GB, from any chip.

The M1 Max with more or less memory:

64 GB39 run well

Neighbouring chips:

Head to head:

Compare it with any Mac or GPU · The M1 Max for local AI · The best Mac for local LLMs · Every Mac

Questions

What is the best LLM for an M1 Max with 32 GB?: Of the models with a published Intelligence Index, Qwen3.8 27B scores highest among those that run well here: 4-bit, about 12–21 tokens a second. The biggest that runs well is Qwen3.6 35B A3B.
Can an M1 Max with 32 GB run Llama 3.3 70B?: Yes, but slowly: 2-bit (lower quality), at about 8–15 tokens a second.
How fast is gpt-oss-20b on an M1 Max?: About 95–168 tokens a second (native 4.25-bit), from the chip’s 400 GB/s memory bandwidth.

Run them with MLX Harness (free, open source) or Apple’s mlx-lm in Terminal: step by step.

How it’s worked out

Memory: the model’s parameters at 8, 6, 4, 3 or 2 bits per weight plus MLX’s group scales (half a bit per weight), or its native format, with room for the context on top. By default macOS lets the GPU use about 75% of unified memory. Speed: memory bandwidth divided by the bytes read per token is the ceiling for one conversation; in llama.cpp’s public Apple Silicon benchmarks real runs reach about 45% to 80% of it, which is the range shown. These are estimates, not measurements: measure your own Mac on MLX Bench. Model figures from each official model card, chip figures from Apple’s Tech Specs. Checked 2026-10-19. Same data as mlx.app/compare.

What can an M1 Max with 32 GB run?