Benchmark result

gpt-oss:20b on NVIDIA GeForce RTX 5070 +1 more — 15.4 tok/s

Measured with Ollama 0.12.9 on August 7, 2026.

What the model built

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 2 runs of this model on NVIDIA GeForce RTX 5070 +1 more · Multi-GPU · median 25.1 tok/s.

sort

134.8tok/s72.8LM Studio
215.4tok/s68.2Ollama · MXFP4this run

Compare the nearest 5 →

gpt-oss:20b on other hardware

Reproduce this run

toolOllama 0.12.9

modelgpt-oss:20b (MXFP4)

context65,536

clientv0.4.30+97

Set those in Ollama, then:

llm-benchmark benchmark --model "gpt-oss:20b" --tool "Ollama"

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.