LLM Inference Benchmark Explorer

DeepSeek V4 Flash 0731 on 2× DGX Spark — NVFP4, vLLM, TP2, speculative decoding k=5 inference benchmark

Open this configuration in the LLM Inference Benchmark Explorer

ModelDeepSeek V4 Flash 0731
Parameters304B
Intelligence Index34.3
Agentic Index41
Device2× DGX Spark
QuantizationNVFP4
Max C4
Chat Capacity16
Agentic Capacity6
TP2
DP—
PP—
EnginevLLM
Speculative DecodingYes
CTTFT (ms)TPS (tok/s)Status
130847.34PASS
241632.37PASS
453222.53PASS
868615.65FAIL

In OpenZeka's measurement, DeepSeek V4 Flash 0731 (304B parameters), served in NVFP4 format with vLLM and speculative decoding on 2× DGX Spark (TP=2), reached a generation speed of 47.3 tok/s per request and a time to first token (TTFT) of 308 ms with a single request.

Considering the speed targets and the available KV cache capacity, the estimated capacity is 16 users for chat use and 6 for agentic use. The realistic capacity will likely fall between these two values. In scenarios dominated by coding, tool use, long workflows and multi-agent use, capacity approaches the agentic estimate; where shorter interactions, standard conversations and lighter tasks dominate, it approaches the chat estimate.

Notes: B12X, DSpark k=5, GMU 0.85, block-size 256, KV fp8

Intelligence Index and Agentic Index values are published by Artificial Analysis and are reproduced here with attribution.