Apple · M3 · unified memory

What runs on a MacBook Pro M3 Max 48GB

48 GB at 400 GB/s. The largest model that fits at Q4_K_M is Mixtral 8x7B, at roughly 25.1 tokens per second. Unified memory means the operating system takes a share, so budget around three quarters of the total.

Runs comfortably

6.1 GB of 36.0 GB · 17%
036 GB
Weights 4.6 GB
KV cache 1.0 GB
Runtime overhead 0.5 GB

Llama 3.1 8B at Q4_K_M leaves 29.9 GB spare on a MacBook Pro M3 Max 48GB. There is room to raise the context length or move up a quantisation level.

Generation44.8tok/s
Prompt processing732tok/s
Max context128Ktokens
KV per 1K tokens0GB

Every quantisation of Llama 3.1 8B on a MacBook Pro M3 Max 48GB

Highlighted row is the highest quality that still fits at 8K context.

QuantisationbpwWeights TotalVerdictMax ctxtok/s
FP16 / BF16 16.00 15.0 GB 16.5 GB Runs comfortably 128K 14.9
INT8 / W8A8 8.50 7.7 GB 9.2 GB Runs comfortably 128K 28.0
Q8_0 (GGUF) 8.50 7.7 GB 9.2 GB Runs comfortably 128K 28.0
FP8 (E4M3) 8.00 7.5 GB 9.0 GB Runs comfortably 128K 28.8
Q6_K 6.56 6.1 GB 7.7 GB Runs comfortably 128K 34.6
AWQ 4-bit 4.25 5.4 GB 6.9 GB Runs comfortably 128K 38.8
GPTQ 4-bit 4.25 5.4 GB 6.9 GB Runs comfortably 128K 38.8
MXFP4 4.25 5.4 GB 6.9 GB Runs comfortably 128K 38.8
Q5_K_M 5.67 5.3 GB 6.8 GB Runs comfortably 128K 39.5
Q5_K_S 5.52 5.2 GB 6.7 GB Runs comfortably 128K 40.5
Q4_K_M 4.85 4.6 GB 6.1 GB Runs comfortably 128K 44.8
Q4_K_S 4.58 4.4 GB 5.9 GB Runs comfortably 128K 46.8
Q4_0 4.55 4.4 GB 5.9 GB Runs comfortably 128K 47.0
IQ4_XS 4.25 4.1 GB 5.6 GB Runs comfortably 128K 49.5
Q3_K_M 3.91 3.8 GB 5.4 GB Runs comfortably 128K 52.6
IQ3_M 3.70 3.7 GB 5.2 GB Runs comfortably 128K 54.7
IQ3_XXS 3.06 3.2 GB 4.7 GB Runs comfortably 128K 62.4
Q2_K 2.63 2.8 GB 4.3 GB Runs comfortably 128K 68.9
IQ2_XXS 2.06 2.3 GB 3.9 GB Runs comfortably 128K 79.9
IQ1_M 1.75 2.1 GB 3.6 GB Runs comfortably 128K 87.6

Models on a MacBook Pro M3 Max 48GB

Q4_K_M at 8K context. Bandwidth sets the speed; capacity sets the ceiling.

ModelParamsWeightsVerdictMax ctxtok/s
DeepSeek-R1 671B-A37B 671B 379.0 GB Won't fit — 1.85
Qwen3 235B-A22B 235.1B 132.8 GB Won't fit — 3.42
gpt-oss 120B-A5.1B 116.8B 66.0 GB Won't fit — 19.8
Llama 4 Scout 109B-A17B 109B 61.7 GB Won't fit — 6.34
Qwen2.5 72B 72.7B 41.2 GB Won't fit — 2.56
Llama 3.1 70B 70.5B 40.0 GB Won't fit — 2.82
Mixtral 8x7B 46.7B 26.4 GB Runs comfortably 32K 25.1
Command R 35B 35.0B 20.1 GB Runs comfortably 12K 9.20
Yi-1.5 34B 34.4B 19.5 GB Runs comfortably 32K 11.3
Qwen3 32B 32.8B 18.6 GB Runs comfortably 67K 11.7
Qwen2.5-Coder 32B 32.8B 18.6 GB Runs comfortably 67K 11.7
DeepSeek-R1-Distill-Qwen 32B 32.8B 18.6 GB Runs comfortably 67K 11.7
Qwen3 30B-A3B 30.5B 17.3 GB Runs comfortably 128K 64.6
Gemma 3 27B 27.4B 15.6 GB Runs comfortably 128K 14.1
Mistral Small 24B 23.6B 13.4 GB Runs comfortably 32K 16.4
gpt-oss 20B-A3.6B 20.9B 11.9 GB Runs comfortably 128K 81.5
Qwen3 14B 14.8B 8.5 GB Runs comfortably 128K 25.3
Phi-4 14B 14.7B 8.4 GB Runs comfortably 16K 25.1
Mistral NeMo 12B 12.3B 7.0 GB Runs comfortably 128K 30.0
Gemma 3 12B 12.2B 7.0 GB Runs comfortably 128K 30.4
GLM-4 9B 9.4B 5.4 GB Runs comfortably 128K 41.2
Qwen3 8B 8.2B 4.7 GB Runs comfortably 128K 43.4
Llama 3.1 8B 8.0B 4.6 GB Runs comfortably 128K 44.8
DeepSeek-R1-Distill-Llama 8B 8.0B 4.6 GB Runs comfortably 128K 44.8
Qwen2.5 7B 7.6B 4.4 GB Runs comfortably 128K 49.8
Mistral 7B v0.3 7.3B 4.1 GB Runs comfortably 32K 49.6
Gemma 3 4B 4.3B 2.5 GB Runs comfortably 128K 84.7
Qwen3 4B 4.0B 2.3 GB Runs comfortably 128K 79.4
Llama 3.2 3B 3.2B 1.8 GB Runs comfortably 128K 99.7
Llama 3.2 1B 1.2B 0.7 GB Runs comfortably 128K 267

Direct answers

Llama 3.1 8B on MacBook Pro M3 Max 48GB
8.0BLlama 3.1 70B on MacBook Pro M3 Max 48GB
70.5BLlama 3.2 3B on MacBook Pro M3 Max 48GB
3.2BLlama 3.2 1B on MacBook Pro M3 Max 48GB
1.2BLlama 4 Scout 109B-A17B on MacBook Pro M3 Max 48GB
109BQwen3 8B on MacBook Pro M3 Max 48GB
8.2BQwen3 14B on MacBook Pro M3 Max 48GB
14.8BQwen3 32B on MacBook Pro M3 Max 48GB
32.8B