Benchmark result
Qwen3.8-27B-Q4_K_S on AMD Radeon RX 6700/6700 XT/6750 XT / 6800M/6850M XT +1 more — 31.2 tok/s
Measured with llama.cpp b2-76101c9 on September 24, 2026.
What the model built
Open full screen →The coding scenario asks for a playable game in a single HTML file. This is exactly what the model returned, unedited.
How this run compares
2nd fastest of 2 runs of this model on AMD Radeon RX 6700/6700 XT/6750 XT / 6800M/6850M XT +1 more · Multi-GPU · median 35.5 tok/s.
sort
- 139.8tok/s73.0llama.cpp · Q4_K_S
- 231.2tok/s82.4llama.cpp · Q4_K_Sthis run
Reproduce this run
toolllama.cpp b2-76101c9
modelQwen3.8-27B-Q4_K_S (Q4_K_S)
context65,536
thinkingon
effortxhigh (model default)
samplingtemperature 1 · top_p 0.949999988079071 · top_k 20 · min_p 0 · presence_penalty 0 · repeat_penalty 1
clientv0.4.80+98
Set those in llama.cpp, then:
llm-benchmark benchmark --model "Qwen3.8-27B-Q4_K_S" --tool "llama.cpp"The client prompts for context length and thinking mode, and for the KV cache dtype on Unsloth Studio. Sampling is left at the model default — the values above are what the tool reported using, not overrides the client sent.