Benchmark result
gpt-oss-20b-MXFP4 on 2× NVIDIA GeForce RTX 5060 — 95.6 tok/s
Measured with llama.cpp b11253-86ea01d05 on September 30, 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 2× NVIDIA GeForce RTX 5060 · Multi-GPU · median 119.7 tok/s.
sort
- 1143.7tok/s77.0llama.cpp
- 295.6tok/s74.0llama.cppthis run
Reproduce this run
toolllama.cpp b11253-86ea01d05
modelgpt-oss-20b-MXFP4
context131,072
thinkingon
effortmedium (model default)
samplingtemperature 1 · top_p 0.949999988079071 · top_k 20 · min_p 0.05000000074505806 · presence_penalty 0 · repeat_penalty 1.0499999523162842
clientv0.4.85+101
Set those in llama.cpp, then:
llm-benchmark benchmark --model "gpt-oss-20b-MXFP4" --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.