Pair

Qwen3.8-Flash-Next

Name: OpenWeights Terminal measured pairs
Creator: Dvidia
License: https://owterminal.com/cite

One quoted capture. Rank only lives inside a filter.

RTX PRO 6000 · Q4_XL · llama.cpp b11412

Decode · tok/s

92.8

Prefill

—

tok/100W

15.5

使用GPUは、RTX PRO 6000。Qwen3.8-flash-next-q4xl … 生成 92.82 t/s … GPURAM 81.07 GB。llama.cpp（build 11412）、500文字までの日本語生成。

First-person llama.cpp bench. Rank is the quoted 92.82 generation t/s. Filename quant is q4xl.

Officialclaimed