Benchmark result
kwangsuklee/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-GGUF:latest on Apple M4 Pro — 11.7 tok/s
Measured with Ollama 0.32.5 on July 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; the frame only adds a line that tells this page how tall it is.
Reproduce this run
toolOllama 0.32.5
modelkwangsuklee/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-GGUF:latest (Q4_K_M)
context8,192
clientv0.4.30+97
Set those in Ollama, then:
llm-benchmark benchmark --model "kwangsuklee/Qwen3.5-27B-Claude-4.6-Opus-Reasoning-Distilled-GGUF:latest" --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.