RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can GLM 4.5V run on the DGX Spark?

Yes. At 8K context the best fit is Q6_K, using about 84.4 GB of the 109 GB usable on the DGX Spark. Estimated speed ceiling: ~23 tokens/sec.

Memory by quant and context

Quant4K ctx8K ctx32K ctxMax tok/s (est., 8K)
2-bit (Q2_K)39.0 GB Fits39.7 GB Fits44.1 GB Fits51
3-bit (Q3_K_M)50.4 GB Fits51.2 GB Fits55.5 GB Fits39
4-bit (Q4_K_M) •60.7 GB Fits61.4 GB Fits65.7 GB Fits32
5-bit (Q5_K_M)72.9 GB Fits73.6 GB Fits77.9 GB Fits27
6-bit (Q6_K)83.7 GB Fits84.4 GB Fits88.7 GB Fits23
8-bit (Q8_0) •107 GB Tight108 GB Tight112 GB No18
16-bit (F16/BF16)202 GB No203 GB No207 GB No—

• real GGUF file size from ggml-org/GLM-4.5V-GGUF. Speed is a bandwidth ceiling (273 GB/s), not a benchmark. Methodology.

More

GLM 4.5V 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 (zai-org/GLM-4.5V), NVIDIA specs. Data as of 2026-11-17.