LLM Inference Benchmark Explorer

Qwen3.8-2.4T-A95B on DGX B300 — NVFP4, vLLM, TP8 inference benchmark

Open this configuration in the LLM Inference Benchmark Explorer

ModelQwen3.8-2.4T-A95B
Parameters2.4T
Intelligence Index39.9
Agentic Index50.1
DeviceDGX B300
QuantizationNVFP4
Max C64+
Chat Capacity192
Agentic Capacity71
TP8
DP—
PP—
EnginevLLM
Speculative Decoding—
CTTFT (ms)TPS (tok/s)Status
1113115.87PASS
294104.57PASS
416987.07PASS
819275.53PASS
1636558.98PASS
3242245.51PASS
6471931.30PASS

In OpenZeka's measurement, Qwen3.8-2.4T-A95B (2.4T parameters), served in NVFP4 format with vLLM on DGX B300 (TP=8), reached a generation speed of 115.9 tok/s per request and a time to first token (TTFT) of 113 ms with a single request.

Considering the speed targets and the available KV cache capacity, the estimated capacity is 192 users for chat use and 71 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: Model: Inferact/Qwen3.8-2.4T-A95B-NVFP4, 262K context

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