DGX Spark x2
22recorded runs
474.0best tok/s
22sourced
How it compares — 128 GB per card · multi-GPU
All benchmarks Compare DGX Spark x2 →
| GPU | Model | Quant | Backend | Ctx | tok/s | prompt tok/s | VRAM GB | Source |
|---|---|---|---|---|---|---|---|---|
| DGX Spark x2 · 128 GB | Qwen3.8-Flash-Next-hibrid48 | NVFP4 (output head) | vllm 0.29 | 262144 | 474.0 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | exl3 v1.4.2 | 4000 | 154.0 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.7bpw (per-expert) | vllm upstream vLLM with 21 patches | 240000 | 141.0 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | tensorfold bd0024d | 384 | 101.1 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | Qwen3.8-Flash-Next-hibrid48 | NVFP4 | vllm 0.30 | 1000 | 99.0 ✓ | 2543.0 | — | src |
| DGX Spark x2 · 128 GB | Qwen3.8-Flash-Next | INT4-AutoRound | vllm 0.30.0 with patches (v5.2 image, sha… | 128000 | 93.0 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | Qwen3.8-Flash-Next-hibrid48 | NVFP4 | vllm 0.30 | 128000 | 89.8 ✓ | 3233.0 | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | tensorfold bd0024d | 128000 | 89.2 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | Qwen3.8-Flash-Next-hibrid48 | NVFP4 | vllm 0.30 | 256000 | 81.6 ✓ | 3094.0 | — | src |
| DGX Spark x2 · 128 GB | Qwen3.8-Flash-Next | NVFP4 | vllm vLLM + patches from github.com/bilik… | 262144 | 80.0 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | exl3 v1.4.2 | 2000 | 74.6 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | GLM-5.3-Flash | NVFP4 | vllm 0.28.1rc1.dev580 | 126720 | 46.3 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | exl3 v1.4.2 | 1000 | 45.5 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | GLM-5.3-Flash | W4A16 AutoRound | vllm v0.31.0 | 400000 | 43.1 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | GLM-5.3-Flash | W4A16 AutoRound | vllm v0.31.0 | 1000 | 39.4 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | GLM-5.3-Flash | W4A16 AutoRound | vllm v0.31.0 | 128000 | 38.5 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | exl3 v1.4.2 | 67000 | 35.0 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | exl3 v1.4.2 | 268000 | 32.6 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | exl3 v1.4.2 | 17000 | 32.4 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | DeepSeek-V4.1-Flash | EXL3 2.9bpw | exl3 v1.4.2 | 559000 | 30.5 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | GLM-5.3-Flash | NVFP4 | vllm glm53-sm121-v13 | 327680 | 30.0 ✓ | — | — | src |
| DGX Spark x2 · 128 GB | GLM-5.3-Flash | EXL3 4-bit | vllm Entrpi/vllm-glm-5.3-flash-spark | 3700 | 27.4 ✓ | 1369.0 | — | src |