Benchmark result
Dirk-Qwen3.8-27B-UD-IQ4_XS on 2× NVIDIA GeForce RTX 5060 — 38.1 tok/s
Measured with llama.cpp b10884-434ddbbc0 on September 11, 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; the frame only adds a line that tells this page how tall it is.
How this run compares
2nd fastest of 4 runs of this model on 2× NVIDIA GeForce RTX 5060 · Multi-GPU · median 37.5 tok/s.
sort
- 138.2tok/s87.1llama.cpp
- 238.1tok/s82.6llama.cppthis run
- 337.0tok/s87.8llama.cpp
- 436.6tok/s84.7llama.cpp
Dirk-Qwen3.8-27B-UD-IQ4_XS on other hardware
Reproduce this run
toolllama.cpp b10884-434ddbbc0
modelDirk-Qwen3.8-27B-UD-IQ4_XS
context65,536
thinkingon
samplingtemperature 0.20000000298023224 · top_p 0.8999999761581421 · top_k 0 · min_p 0.05000000074505806 · presence_penalty 0 · repeat_penalty 1
clientv0.4.59+97
Set those in llama.cpp, then:
llm-benchmark benchmark --model "Dirk-Qwen3.8-27B-UD-IQ4_XS" --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.