2x DGX Spark Runs DeepSeek v4.1 Flash at Record 103 tok/s for Code

A user ran DeepSeek v4.1 Flash with vLLM on two NVIDIA DGX Spark nodes using custom quantized weights, hitting a record 103 tok/s on code; an open-source TP2 patch further cut time-to-first-token from 33.8s to 4.9s.

2026-10-12 ~ 2026-10-12 · 2 related posts

• DeepSeek-V4.1-Flash on 2x DGX Spark: TP2 Patches Cut First Token From 33.8s to 4.9s — rez0__ · 2026-10-12
• DeepSeek v4.1 Flash served at 103 tok/s on 2x DGX Spark with vLLM — EAccelerate_42 · 2026-10-12