Machine

RX 6700 6700 Xt 6750 Xt 6800m 6850m Xt · Ryzen 7 5800X #ba736d42

9 runs of 9 models since Sep 29, 2026. Last run Oct 11, 2026.

ShareRedditX
9
runs
9
models
9
configs
0
configs run 2+ times

Hardware

GPU
AMD Radeon RX 6700/6700 XT/6750 XT / 6800M/6850M XT 12 GB
CPU
AMD Ryzen 7 5800X 8-Core Processor 8 cores
RAM
31 GB
OS
CachyOS
Tools
llama.cpp 9
Client
0.4.91
All AMD Radeon RX 6700/6700 XT/6750 XT / 6800M/6850M XT benchmarks

Models on this machine

One dot per model, at its best config. Up is better answers, right is faster. Time to finish counts every token the model wrote, thinking included, so a model that talks a lot ends up further left.

Best trade-off: quality 60+, and nothing else here is both faster and betterOther models · dot size = number of runs
below 20 tok/s: slower than you read4050607080901002050tok/s (log scale) →quality →

Medians of each model's best config, 9 models.

Best config per model

sort

Runs with the same model, quant, tool, context, thinking and KV cache count as one config. Numbers are the median of its runs, and each dot is one run on the 0–100 quality scale. ★ marks the best config with at least 2 rated runs.

Qwen3.8-27B-GSQ-RCO-IQ3_XXS-mtp
1 run · 1 config
★ bestllama.cpp32k ctxthinking oneffort medium
84.5
1 rated run
28.6 tok/s
Tiel-Coder-35B-A3B-MTP-UD-Q4_K_XL
1 run · 1 config
★ bestQ4_Kllama.cpp256k ctxthinking on
82.4
1 rated run
33.9 tok/s
Underdog-Saluki-27B-1.0-IQ2-mix-MTP
1 run · 1 config
★ bestllama.cpp72k ctxthinking oneffort medium
82.2
1 rated run
30.8 tok/s
Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_XXS-mtp
1 run · 1 config
★ bestllama.cpp32k ctxthinking oneffort medium
81.4
1 rated run
28.6 tok/s
Qwen3.8-35B-A3B-Q4_K_M
1 run · 1 config
★ bestQ4_K_Mllama.cpp256k ctxthinking on
76.6
1 rated run
33.4 tok/s
Shadow-Siren-26B-A4B.i1-Q4_K_M
1 run · 1 config
★ bestQ4_K_Mllama.cpp64k ctxthinking on
72.4
1 rated run
42.4 tok/s
Melody1437-12B.i1-Q4_K_M
1 run · 1 config
★ bestQ4_K_Mllama.cpp256k ctxthinking default
71.9
1 rated run
44.6 tok/s
MiniCPM5-2B-F16
1 run · 1 config
★ bestllama.cpp128k ctxthinking on
56.0
1 rated run
43.0 tok/s
Dark-Oddity-12B-v2.0-Q5_K_M
1 run · 1 config
★ bestQ5_K_Mllama.cpp32k ctxthinking on
43.5
1 rated run
18.6 tok/s

Compare with another machine

Run history