RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can SmolLM2 360M Instruct run on the DGX Spark?

Yes. At 8K context the best fit is F16, using about 1.5 GB of the 109 GB usable on the DGX Spark. Estimated speed ceiling: ~376 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctxMax tok/s (est., 8K)
2-bit (Q2_K) •0.9 GB Fits1.0 GB Fits1248
3-bit (Q3_K_M) •0.9 GB Fits1.0 GB Fits1163
4-bit (Q4_K_M) •0.9 GB Fits1.1 GB Fits1009
5-bit (Q5_K_M) •0.9 GB Fits1.1 GB Fits942
6-bit (Q6_K) •1.0 GB Fits1.1 GB Fits743
8-bit (Q8_0) •1.0 GB Fits1.2 GB Fits707
16-bit (F16/BF16) •1.3 GB Fits1.5 GB Fits376

• real GGUF file size from unsloth/SmolLM2-360M-Instruct-GGUF. Speed is a bandwidth ceiling (273 GB/s), not a benchmark. Methodology.

More

SmolLM2 360M Instruct on every GPU and Mac · Every model that runs on the DGX Spark · Other models that fit 64 GB VRAM · Try other settings in the calculator

Data: Hugging Face Hub (HuggingFaceTB/SmolLM2-360M-Instruct), NVIDIA specs. Data as of 2026-11-17.