Benchmark result
unsloth/gpt-oss-20b-GGUF on Intel(R) Arc(TM) 140V GPU (16GB) — 21.3 tok/s
Measured with llama.cpp unsloth-studio/v0.1.810-beta on September 18, 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 3 runs of this model on Intel(R) Arc(TM) 140V GPU (16GB) · median 21.3 tok/s.
sort
- 123.0tok/s59.2llama.cpp
- 221.3tok/s60.7llama.cppthis run
- 320.9tok/s28.4llama.cpp
unsloth/gpt-oss-20b-GGUF on other hardware
Reproduce this run
toolllama.cpp unsloth-studio/v0.1.810-beta
modelunsloth/gpt-oss-20b-GGUF
context8,192
effortmedium (model default)
samplingtemperature 0.800000011920929 · top_p 0.949999988079071 · top_k 40 · min_p 0.05000000074505806 · presence_penalty 0 · repeat_penalty 1
clientv0.4.59+97
Set those in llama.cpp, then:
llm-benchmark benchmark --model "unsloth/gpt-oss-20b-GGUF" --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.