Benchmark result
Qwen3.8-9B-Q4_K_M on NVIDIA GeForce RTX 4060 Laptop GPU — 34.1 tok/s
Measured with llama.cpp b11985-6d194531e on October 30, 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
4th fastest of 5 runs of this model on NVIDIA GeForce RTX 4060 Laptop GPU · median 34.2 tok/s.
sort
- 135.7tok/s71.1llama.cpp · Q4_K_M
- 234.6tok/s71.4llama.cpp · Q4_K_M
- 334.2tok/s70.6llama.cpp · Q4_K_M
- 434.1tok/s72.9llama.cpp · Q4_K_Mthis run
- 533.8tok/s77.7llama.cpp · Q4_K_M
Qwen3.8-9B-Q4_K_M on other hardware
Reproduce this run
toolllama.cpp b11985-6d194531e
modelQwen3.8-9B-Q4_K_M (Q4_K_M)
context32,768
think limitnone
samplingtemperature 0.800000011920929 · top_p 0.949999988079071 · top_k 40 · min_p 0.05000000074505806 · presence_penalty 0 · repeat_penalty 1
clientv0.4.98+102
Set those in llama.cpp, then:
llm-benchmark benchmark --model "Qwen3.8-9B-Q4_K_M" --tool "llama.cpp"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.