Benchmark result
theprint-12b-moe-3a-q4_k_m on NVIDIA GeForce RTX 4060 Laptop GPU — 34.6 tok/s
Measured with llama.cpp b11985-6d194531e on October 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.
How this run compares
2nd fastest of 10 runs of this model on NVIDIA GeForce RTX 4060 Laptop GPU · median 34.0 tok/s.
sort
- 134.7tok/s40.0llama.cpp · Q4_K_M
- 234.6tok/s42.5llama.cpp · Q4_K_Mthis run
- 334.5tok/s31.1llama.cpp · Q4_K_M
- 434.1tok/s32.6llama.cpp · Q4_K_M
- 534.1tok/s27.8llama.cpp · Q4_K_M
- 634.0tok/s38.7llama.cpp · Q4_K_M
- 734.0tok/s30.0llama.cpp · Q4_K_M
- 834.0tok/s28.0llama.cpp · Q4_K_M
- 933.5tok/s31.9llama.cpp · Q4_K_M
- 1033.2tok/s31.5llama.cpp · Q4_K_M
Reproduce this run
toolllama.cpp b11985-6d194531e
modeltheprint-12b-moe-3a-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 "theprint-12b-moe-3a-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.