llamaperf

DeepSeek V4.1

on NVIDIA RTX 3090

Tone: mixedNVIDIA hardware
Nov 11, 2026
Throughput
2-3 t/s gen
System RAM
32 GB

Summary

User reports running DeepSeek 4.1 on a single RTX 3090 in a 32GB machine at 2-3 t/s. No engine, quant or context length is stated; the user notes it is only a 2-bit quant and that the run is not really useful at the moment. The user plans to add 128GB of RAM, a spare RTX 3060 and possibly another 3090.