DGX Spark x2 vs DGX Spark x3

Best sourced output tok/s per model (single-stream generation) — engine and version, quantization, context and a source link on every row. Hover a value for the quant.

DGX Spark x222 sourced runs · best 474.0 tok/s

DGX Spark x32 verified runs · best 152.9 tok/s

Specs and evidence

	DGX Spark x2	DGX Spark x3
VRAM	128 GB	128 GB
Sourced runs on the board	22	2
Models measured	4	2
Best output tok/s	474.0 tok/s	152.9 tok/s
Highest measured peak VRAM	—	—
Street price (latest snapshot)	—	—

Sourced rows only. Peak VRAM is recorded by the person who ran the benchmark, so it is a floor for what the card can hold, not a specification. Price snapshots.

Head-to-head

Only one model has been measured on both cards (DeepSeek-V4.1-Flash). On that single data point DGX Spark x2 records 154.0 tok/s and DGX Spark x3 records 46.0 tok/s — treat it as one data point, not a ranking.

The measured matrix

Model	DGX Spark x2	DGX Spark x3
Qwen3.8-Flash-Next-hibrid48	474.0 NVFP4 (output head)	—
DeepSeek-V4.1-Flash	154.0 EXL3 2.9bpw	46.0 EXL3 3.5bpw
DeepSeek-V4.1-Flash-1017	—	152.9 EXL3 3.5bpw
Qwen3.8-Flash-Next	93.0 INT4-AutoRound	—
GLM-5.3-Flash	46.3 NVFP4	—

Head-to-head (models tested on both): DGX Spark x2 1 · DGX Spark x3 0 · tied 0 · data as of Oct 17, 2026. Full rules: methodology.

Related reading: RTX 5090 vs 3090 · quantization · Value index · Full board