Benchmark result

gpt-oss-20b-MLX-8bit on Apple M5 Max — 95.7 tok/s

Measured with oMLX 0.31.4.dev132+g94cdcae13 on October 10, 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.

Reproduce this run

tooloMLX 0.31.4.dev132+g94cdcae13

modelgpt-oss-20b-MLX-8bit (8bit)

context131,072

kv cachenone

thinkingon

effortmedium (model default)

think limitnone

samplingtemperature 1 · top_p 0.95 · top_k 20 · repetition_penalty 1

clientv0.4.85+101

Set those in oMLX, then:

llm-benchmark benchmark --model "gpt-oss-20b-MLX-8bit" --tool "oMLX"

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.