RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can DeepSeek V3.2 run on the Mac M3 Ultra 512GB?

Yes. At 8K context the best fit is Q3_K_M, using about 300 GB of the 384 GB usable on the Mac M3 Ultra 512GB. Estimated speed ceiling: ~43 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctx128K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •229 GB Fits230 GB Fits232 GB Fits240 GB Fits56
3-bit (Q3_K_M) •299 GB Fits300 GB Fits302 GB Fits309 GB Fits43
4-bit (Q4_K_M) •379 GB Tight379 GB Tight381 GB Tight389 GB No34
5-bit (Q5_K_M) •444 GB No445 GB No447 GB No455 GB No—
6-bit (Q6_K) •514 GB No515 GB No517 GB No525 GB No—
8-bit (Q8_0) •665 GB No666 GB No668 GB No675 GB No—
16-bit (F16/BF16) •1251 GB No1251 GB No1253 GB No1261 GB No—

• real GGUF file size from unsloth/DeepSeek-V3.2-GGUF. Speed is a bandwidth ceiling (819 GB/s), not a benchmark. Methodology.

More

DeepSeek V3.2 on every GPU and Mac · Every model that runs on the Mac M3 Ultra 512GB · Apple Silicon guide · Try other settings in the calculator

Data: Hugging Face Hub (deepseek-ai/DeepSeek-V3.2), Apple specs. Data as of 2026-11-17.