RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can MiniMax M2.7 run on the RTX 3080 12GB?

Not fully in VRAM. Even at 2-bit with 8K context, MiniMax M2.7 needs about 82.6 GB, more than the 11.4 GB usable on the RTX 3080 12GB. It can still run with offload (slower): Q3_K_M with part of the model in system RAM, about 7.7–16 tokens/sec (estimate). Details and command.

Runs with offload (slower)

Runs with offload 3-bit (Q3_K_M) at 8K context: all attention layers plus the experts of 4 of 62 MoE layers on the RTX 3080 12GB (about 9.8 GB), and the experts of the other 58 layers in system RAM (about 86.4 GB). Needs a PC with at least 128 GB of system RAM. Estimated speed: 7.7–16 tokens/sec (slow but usable), versus a bandwidth ceiling of ~187 tok/s if it fit fully in VRAM.

llama-server -hf unsloth/MiniMax-M2.7-GGUF:Q3_K_M -c 8192 -ngl 999 --n-cpu-moe 58

Estimate, not a benchmark. The command downloads the GGUF from Hugging Face. The speed range spans dual-channel DDR4-3200 to DDR5-6000 system RAM; your CPU, RAM speed and llama.cpp version move it. If llama.cpp runs out of GPU memory, raise --n-cpu-moe by 1-2; if VRAM is left over, lower it for more speed. Recent llama.cpp builds can also pick the split automatically. How we estimate offload.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K)81.6 GB No82.6 GB No88.4 GB No112 GB No—
3-bit (Q3_K_M) •96.0 GB No97.0 GB No103 GB No126 GB No—
4-bit (Q4_K_M) •132 GB No133 GB No139 GB No162 GB No—
5-bit (Q5_K_M) •159 GB No160 GB No166 GB No189 GB No—
6-bit (Q6_K) •177 GB No178 GB No184 GB No207 GB No—
8-bit (Q8_0) •228 GB No229 GB No235 GB No258 GB No—
16-bit (F16/BF16) •428 GB No429 GB No435 GB No458 GB No—

• real GGUF file size from unsloth/MiniMax-M2.7-GGUF. Speed is a bandwidth ceiling (912 GB/s), not a benchmark. Methodology.

Hardware that does fit it

RTX PRO 6000 BlackwellDGX SparkRyzen AI Max+ 395 128GBMac M3 Max 128GBMac M4 Max 128GBH200

What fits instead

"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.

More

MiniMax M2.7 on every GPU and Mac · Every model that runs on the RTX 3080 12GB · Other models that fit 12 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (MiniMaxAI/MiniMax-M2.7), NVIDIA specs. Model data last checked 2026-11-17; refreshed hourly.