Serving DeepSeek-V4-Flash-Vision on 4× DGX Spark: images, DSpark and DCP in one stack

DeepSeek-V4-Flash-Vision-Exp on four GB10 nodes, ported by hand onto the sparkinfer B12X stack before upstream support landed. The interesting part was not the vision tower — it was making DSpark speculative decoding and decode-context-parallel run together, which the stock stack refuses. Decode kept, KV doubled, images cost about a second.

• DeepSeek-V4-Flash
• vLLM
• multi-node
• TP4
• DCP
• DGX Spark
• FP8
• MoE
• MLA
• speculative decoding
• thinking mode
• tool calling
• long context
• multimodal

• time~30 min per node, plus ~10 min per boot
• yields4-node cluster serving DeepSeek-V4-Flash-Vision-Exp at 66–71 tok/s decode, 393K context, 3.5M-token KV pool, images in and thinking on