Benchmark result
qwen3.8:27b-mlx on Apple M1 Max — 20.1 tok/s
Measured with Ollama 0.40.15 on November 15, 2026.
Support llm-bench.io
llm-bench.io is free. If it helped you, I'd appreciate your support to cover the infrastructure costs.
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 2 runs of this model on Apple M1 Max · median 21.1 tok/s.
sort
- 122.1tok/s84.1Ollama · nvfp4
- 220.1tok/s87.5Ollama · nvfp4this run
Reproduce this run
toolOllama 0.40.15
modelqwen3.8:27b-mlx (nvfp4)
context65,536
thinkingon
samplingtemperature 1 · top_p 0.95 · top_k 20 · min_p 0 · presence_penalty 0 · repeat_penalty 1
clientv0.4.99+102
Set those in Ollama, then:
llm-benchmark benchmark --model "qwen3.8:27b-mlx" --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.