Machine

RX 6700 6700 Xt 6750 Xt 6800m 6850m Xt · Ryzen 5 5600 #aed0cdfe

6 runs of 5 models since Aug 24, 2026. Last run Aug 24, 2026.

ShareRedditX
6
runs
5
models
5
configs
1
configs run 2+ times

Hardware

GPU
AMD Radeon RX 6700/6700 XT/6750 XT / 6800M/6850M XT 11 GB
CPU
AMD Ryzen 5 5600 6-Core Processor 6 cores
RAM
15 GB
OS
CachyOS
Tools
LM Studio 5 · vLLM 1
Client
0.4.40
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 read0204060801002050100tok/s (log scale) →quality →outliers sit at the edges

Medians of each model's best config, 5 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.

Ornith-1.5-9B-MTP-Q6_K
1 run · 1 config
★ bestQ6_KLM Studio8k ctxthinking default
76.7
1 rated run
44.4 tok/s
Ornith-1.5-9B-MTP-Q6_0_ROCMFPX_AGENT
1 run · 1 config
★ bestQ6_0LM Studio8k ctxthinking default
66.4
1 rated run
34.2 tok/s
Ornith-1.5-9B-MTP-Q4_0_ROCMFP4_COHERENT
2 runs · 1 config
★ bestQ4_0LM Studio8k ctxthinking default
×2 runs
57.757.0–58.4
steady
58.3 tok/s
57.3–59.3
Ornith-1.5-9B-MTP-IQ4_XS
1 run · 1 config
★ bestLM Studio8k ctxthinking default
56.1
1 rated run
50.8 tok/s
Qwen3.8-27B-ROCMFP2
1 run · 1 config
★ bestvLLM64k ctxthinking on
0.0
1 rated run
5167 tok/s

Compare with another machine

Run history