Benchmark result

Qwen3.8-Flash-Next-UD-IQ3_XXS on Apple M5 Ultra — 116.3 tok/s

Measured with oMLX 0.7.3 on November 6, 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 Apple M5 Ultra · median 128.6 tok/s.

sort

1140.9tok/s76.7oMLX · KV none
2116.3tok/s75.3oMLX · KV nonethis run

Compare the nearest 5 →

Reproduce this run

tooloMLX 0.7.3

modelQwen3.8-Flash-Next-UD-IQ3_XXS

context65,536

kv cachenone

effortxhigh (model default)

think limitnone

samplingtemperature 1 · top_p 0.95 · top_k 20 · min_p 0 · presence_penalty 1.5 · repetition_penalty 1

clientv0.4.98+102

Set those in oMLX, then:

llm-benchmark benchmark --model "Qwen3.8-Flash-Next-UD-IQ3_XXS" --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.