readygpu

Mac mini M4 Pro 48GB vs NVIDIA GeForce RTX 5080

For running models locally, priced in the US. At 32k context the first runs 58 models fully in memory, the second 34.

In short

The Mac mini M4 Pro 48GB has more memory for models: 48 GB against 16 GB. At 32k context the Mac mini M4 Pro 48GB runs 58 of the models we track fully in memory and the NVIDIA GeForce RTX 5080 34; only the Mac mini M4 Pro 48GB holds Qwen 3.8 27B. On the models both run fully in memory, the NVIDIA GeForce RTX 5080 is about 3.4× as fast as the Mac mini M4 Pro 48GB (estimated, median across models). In the US the Mac mini M4 Pro 48GB costs $2,149 at eBay and the NVIDIA GeForce RTX 5080 $1,599 at eBay: the NVIDIA GeForce RTX 5080 is $550 cheaper.

Updated . Speeds are estimates; how they are worked out.

Mac mini M4 Pro 48GB

$2,149

eBay

Add to build
NVIDIA GeForce RTX 5080

$1,599

eBay

Add to build

Specifications

SpecMac mini M4 Pro 48GBNVIDIA GeForce RTX 5080
ChipApple M4 Pro (20-core GPU)–
Memory bus256-bit256-bit
Memory bandwidth273 GB/s960 GB/s
FP16 compute8.08 TFLOPS113 TFLOPS
SoftwareMetalCUDA
Unified memory48 GB–
GPU can use by default36 GB–
GPU can use, tuned40 GB–
Operating systemmacos–
ProcessorM4 Pro 14-core–
Storage512 GB–
GPU–GeForce RTX 5080
Memory–16 GB GDDR7
Board power–360 W
PCIe generation–Gen 5
PCIe lanes–x16

What each runs

Estimated decode speed with the most precise quant that fits each one est.

ModelMac mini M4 Pro 48GBNVIDIA GeForce RTX 5080
Qwen 3.8 27B6tok/sQ8_0offload
Qwen 3.6 27B6tok/sQ8_0offload
Gemma 4 31B5tok/sQ8_0offload
Qwen 3.6 35B-A3B28tok/sQ8_0offload
Qwen 3.5 27B6tok/sQ8_0offload
Gemma 4 26B-A4B21tok/sQ8_0offload
Qwen 3.5 35B-A3B28tok/sQ8_0offload
GLM 4.7 Flash15tok/sQ8_0offload
Devstral Small 2 24B6tok/sQ8_0offload
Nemotron 3.5 Lightning 30B-A3B30tok/sUD-Q6_K_XLoffload
Kimi Linear 48B-A3B44tok/sQ5_K_Moffload
gpt-oss 20B33tok/sMXFP4110tok/sMXFP4
Qwen3 Coder 30B-A3B17tok/sQ8_0offload
Qwen3 32B5tok/sQ6_Koffload
Gemma 4 12B13tok/sQ8_044tok/sQ8_0
Qwen 3.5 9B17tok/sQ8_056tok/sQ8_0
Mistral Small 3.2 24B6tok/sQ8_0offload
Gemma 3 27B6tok/sQ8_0offload
DeepSeek R1 Distill Qwen 32B5tok/sQ6_Koffload
Ministral 3 14B9tok/sQ8_040tok/sQ5_K_M
Qwen3 14B8tok/sQ8_041tok/sQ4_K_M
DeepSeek R1 0528 Qwen3 8B13tok/sQ8_044tok/sQ8_0
Qwen3 8B13tok/sQ8_044tok/sQ8_0
Phi-4 14B8tok/sQ8_0offload
Qwen 3.5 4B32tok/sQ8_0106tok/sQ8_0
Ministral 3 8B13tok/sQ8_044tok/sQ8_0
Granite 4.2 8B12tok/sQ8_041tok/sQ8_0
Gemma 4 E4B19tok/sQ8_065tok/sQ8_0
LFM2.5 8B-A1B40tok/sQ8_0134tok/sQ8_0
Llama 3.1 8B16tok/sQ6_K55tok/sQ6_K
Qwen2.5 0.5B190tok/sQ8_0637tok/sQ8_0
Qwen3.5 0.8B144tok/sQ8_0483tok/sQ8_0
Qwen3.5 2B96tok/sQ5_K_M321tok/sQ5_K_M
DeepSeek R1 Distill Qwen 1.5B63tok/sQ8_0210tok/sQ8_0
MiniCPM5 2B43tok/sQ8_0146tok/sQ8_0
Spark X2.5 4B41tok/smlx-4bit138tok/smlx-4bit
LFM2.5 2.6B36tok/sQ8_0122tok/sQ8_0
LFM2.5 VL 3B36tok/sQ8_0122tok/sQ8_0
Mellum2.1 12B A2.5B Thinking34tok/sQ8_0113tok/sQ8_0
FrogNano 4B 260931tok/sQ8_0104tok/sQ8_0
granite vision 4.1 4b28tok/sQ8_094tok/sQ8_0
granite 4.2 3b27tok/sQ8_091tok/sQ8_0
Kimi VL A3B Thinking 250626tok/sQ8_0121tok/sQ4_K_M
Kimi VL A3B25tok/smlx-8bit101tok/smlx-6bit
LFM2.5 24B A2B22tok/sQ8_0offload
GLM 4 9B 041419tok/smlx-5bit62tok/smlx-5bit
DeepSeek R1 Distill Qwen 7B18tok/sQ8_060tok/sQ8_0
GLM 4.6V Flash16tok/sQ8_053tok/sQ8_0
DeepSeek R1 Distill Llama 8B14tok/sQ8_046tok/sQ8_0
GLM 4.1V 9B Thinking14tok/smlx-8bit45tok/smlx-8bit
Ministral 3 8B Reasoning 251213tok/sQ8_044tok/sQ8_0
DeepSeek R1 Distill Qwen 14B8tok/sQ8_039tok/sQ4_K_M
Qwen3.8 27B OBLITERATED8tok/smlx-4bitoffload
Devstral Small 25076tok/sQ8_0offload
Magistral Small 25066tok/sQ8_0offload
Magistral Small 25076tok/sQ8_0offload
Magistral Small 25096tok/sQ8_0offload
granite 4.2 30b5tok/sQ6_Koffload
Qwen 3.5 122B-A10Bwon't runoffload
Mistral Small 4 119Bwon't runoffload
DeepSeek V4 Flashwon't runwon't run
GLM 5.3 Flashwon't runwon't run
Qwen 3.8 Flash Nextwon't runwon't run
MiniMax M2.7won't runwon't run
Nemotron 3 Super 120B-A12Bwon't runwon't run
Qwen3 Coder Nextwon't runwon't run
gpt-oss 120Bwon't runwon't run
GLM 4.5 Airwon't runwon't run
Qwen3 Next 80B-A3Bwon't runwon't run
Llama 3.3 70Bwon't runwon't run
DeepSeek R1won't runwon't run
DeepSeek R1 0528won't runwon't run
DeepSeek R1 Distill Llama 70Bwon't runwon't run
DeepSeek R1 Zerowon't runwon't run
DeepSeek V3 0324won't runwon't run
DeepSeek V3.1won't runwon't run
DeepSeek V3.1 Terminuswon't runwon't run
DeepSeek V3.2won't runwon't run
DeepSeek V4 Flash 0731won't runwon't run
DeepSeek V4 Flash Vision Expwon't runwon't run
Devstral 2 123B Instruct 2512won't runwon't run
GLM 4.5Vwon't runwon't run
GLM 4.6won't runwon't run
GLM 4.6Vwon't runwon't run
GLM 4.7won't runwon't run
GLM 5won't runwon't run
GLM 5.1won't runwon't run
GLM 5.2won't runwon't run
GLM 5.3won't runwon't run
Kimi Dev 72Bwon't runwon't run
Kimi K2won't runwon't run
Kimi K2 Instruct 0905won't runwon't run
Kimi K2 Thinkingwon't runwon't run
Kimi K2.5won't runwon't run
MiniMax M2won't runwon't run
MiniMax M2.1won't runwon't run
MiniMax M2.5won't runwon't run
MiniMax M3won't runwon't run
Mistral Medium 3.5 128Bwon't runwon't run

Questions

Which is better for local AI, the Mac mini M4 Pro 48GB or the NVIDIA GeForce RTX 5080?
Choose the Mac mini M4 Pro 48GB to run bigger models. Choose the NVIDIA GeForce RTX 5080 for speed. Choose the NVIDIA GeForce RTX 5080 to spend less.
Which can run bigger models, the Mac mini M4 Pro 48GB or the NVIDIA GeForce RTX 5080?
At 32k context the Mac mini M4 Pro 48GB runs 58 of the models we track fully in memory and the NVIDIA GeForce RTX 5080 34; only the Mac mini M4 Pro 48GB holds Qwen 3.8 27B.
Which is faster for local AI, the Mac mini M4 Pro 48GB or the NVIDIA GeForce RTX 5080?
On the models both run fully in memory, the NVIDIA GeForce RTX 5080 is about 3.4× as fast as the Mac mini M4 Pro 48GB (estimated, median across models).
What is the price difference between the Mac mini M4 Pro 48GB and the NVIDIA GeForce RTX 5080?
In the US the Mac mini M4 Pro 48GB costs $2,149 at eBay and the NVIDIA GeForce RTX 5080 $1,599 at eBay: the NVIDIA GeForce RTX 5080 is $550 cheaper.

More comparisons