LLM Inference Benchmark Explorer

Qwen3.8-27B on DGX B300 — BF16, vLLM, TP1, speculative decoding inference benchmark

Open this configuration in the LLM Inference Benchmark Explorer

ModelQwen3.8-27B
Parameters27B
Intelligence Index33.7
Agentic Index45.8
DeviceDGX B300
QuantizationBF16
Max C64+
Chat Capacity225
Agentic Capacity77
TP—
DP—
PP—
EnginevLLM
Speculative DecodingYes
CTTFT (ms)TPS (tok/s)Status
177189.39PASS
2112178.57PASS
4101178.89PASS
8130142.25PASS
16155145.56PASS
32201113.25PASS
6422567.25PASS

In this measurement, Qwen3.8-27B (27B parameters), served in BF16 format with vLLM and speculative decoding on DGX B300, reached a generation speed of 189.4 tok/s per request and a time to first token (TTFT) of 77 ms with a single request.

Considering the speed targets and the available KV cache capacity, the estimated capacity is 225 users for chat use and 77 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.