LLM Inference Benchmark Explorer

Gemma-4-31B-it on 2× DGX Spark — NVFP4, vLLM, TP2 inference benchmark

Open this configuration in the LLM Inference Benchmark Explorer

ModelGemma-4-31B-it
Parameters31B
Intelligence Index14.7
Agentic Index4.2
Device2× DGX Spark
QuantizationNVFP4
Max C2
Chat Capacity8
Agentic Capacity3
TP2
DP—
PP—
EnginevLLM
Speculative Decoding—
CTTFT (ms)TPS (tok/s)Status
113619.68FAIL
222720.14PASS
424619.76FAIL
828018.81FAIL
1629316.29FAIL
3233713.58FAIL
644209.38FAIL

In this measurement, Gemma-4-31B-it (31B parameters), served in NVFP4 format with vLLM on 2× DGX Spark (TP=2), reached a generation speed of 19.7 tok/s per request and a time to first token (TTFT) of 136 ms with a single request.

Considering the speed targets and the available KV cache capacity, the estimated capacity is 8 users for chat use and 3 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.

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