Machine

2× RX 7800 Xt · Ryzen 7 5800XT #f202dd99

31 runs of 13 models since Sep 18, 2026. Last run Nov 17, 2026.

ShareRedditX
31
runs
13
models
22
configs
7
configs run 2+ times

Hardware

GPU
2× AMD Radeon RX 7800 XT 32 GB
CPU
AMD Ryzen 7 5800XT 8-Core Processor 8 cores
RAM
63 GB
OS
Windows 11 Pro 26H2
Tools
Ollama 31
Client
0.4.89
All 2× AMD Radeon RX 7800 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, 13 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.

aiconjured/Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-Q5_NVFP4
1 run · 1 config
★ bestQ4_K_MOllama64k ctxthinking oneffort xhigh
87.5
1 rated run
28.4 tok/s
logicbeat/qwen3.8-27B_GSQ_RCO
1 run · 1 config
★ bestIQ3_SOllama8k ctxthinking default
83.4
1 rated run
23.9 tok/s
muse-glimmer:30b-q4_K_M
4 runs · 1 config
★ bestQ4_K_MOllama64k ctxthinking oneffort medium
×4 runs
82.878.8–83.6
steady
27.5 tok/s
26.6–28.4
qwen3.8:27b-q4_K_M
6 runs · 5 configs
★ bestQ4_K_MOllama64k ctxthinking oneffort medium
×2 runs
80.879.6–82.0
steady
26.5 tok/s
25.9–27.1
gpt-oss:20b
3 runs · 2 configs
★ bestMXFP4Ollama8k ctxthinking default
×2 runs
72.472.0–72.8
steady
93.4 tok/s
93.2–93.5
gemma4:26b-a4b-it-qat
4 runs · 3 configs
★ bestQ4_0Ollama64k ctxthinking on
×2 runs
70.670.3–70.9
steady
71.7 tok/s
71.3–72.1
gemma4:12b-it-qat
3 runs · 2 configs
★ bestQ4_0Ollama8k ctxthinking default
×2 runs
69.168.4–69.8
steady
54.2 tok/s
54.0–54.5
mannix/ornith-1.5-27b-a3b-coder
2 runs · 1 config
★ bestQ4_K_MOllama64k ctxthinking on
×2 runs
63.463.1–63.8
steady
74.7 tok/s
73.2–76.2
qwen3.5:9b
3 runs · 2 configs
★ bestQ4_K_MOllama8k ctxthinking default
×2 runs
60.453.8–67.0
varies
71.8 tok/s
71.7–71.9
yogeshjog/muse-glimmer-q2
1 run · 1 config
★ bestQ2_KOllama8k ctxthinking default
59.9
1 rated run
30.2 tok/s

Compare with another machine

Run history

Show all 31 runs ▾