deepseek-v4.1IMAGE

DeepSeek V4.1 Flash hits 100+ tok/s on a 12GB RTX 5070

A user reports over 100 tok/s running DeepSeek V4.1 Flash on a 12GB RTX 5070 using the Strata engine's layer-major chunked reading and GSQ-RCO quantization, with more RAM boosting decode speed.

Forha

@ForhanvvArchived 4 days agoOriginally posted 2026.10.26

DeepSeek V4.1 Flash hits 100+ tok/s on a 12GB RTX 5070 (image 1 of 2)

DeepSeek V4.1 Flash hits 100+ tok/s on a 12GB RTX 5070 (image 2 of 2)

Forha @ForhanvvOriginally posted 2026.10.26

Open original on X

Running DeepSeek-V4.1-Flash on a 12GB RTX 5070 at over 100+ tok/s is absolutely absurd. 😭 The Strata engine’s layer-major chunked reading + GSQ-RCO quantization makes consumer MoE offloading feel completely native. Prefill is flying, and more RAM directly boosts decode speed. Local compute is getting ridiculous. 💀

x.com/forhanvv/status/2114429171984437342Archived 2026.10.26 12:05