Benchmark result
m5-qwen3-4b-thinking — 87.9 tok/s
Measured with cliproxyapi manual on September 27, 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.
How this run compares
m5-qwen3-4b-thinking on other hardware
- 89.0tok/s45.6Unknown hardware
- 87.9tok/s51.8Unknown hardwarethis run
Reproduce this run
toolcliproxyapi manual
modelm5-qwen3-4b-thinking
context8,192
clientv0.4.80+98
Set those in cliproxyapi, then:
llm-benchmark benchmark --model "m5-qwen3-4b-thinking" --tool "cliproxyapi"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.