Benchmark result
Llama-3.1-Tulu-3-8B-Q4_K_M on NVIDIA GeForce RTX 4060 Laptop GPU — 42.5 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
4th fastest of 10 runs of this model on NVIDIA GeForce RTX 4060 Laptop GPU · median 42.4 tok/s.
sort
- 143.9tok/s47.3llama.cpp · Q4_K_M
- 243.2tok/s47.6llama.cpp · Q4_K_M
- 343.0tok/s44.5llama.cpp · Q4_K_M
- 442.5tok/s54.1llama.cpp · Q4_K_Mthis run
- 542.4tok/s56.8llama.cpp · Q4_K_M
- 642.4tok/s40.8llama.cpp · Q4_K_M
- 742.2tok/s53.6llama.cpp · Q4_K_M
- 841.7tok/s40.7llama.cpp · Q4_K_M
- 940.9tok/s40.0llama.cpp · Q4_K_M
- 1039.5tok/s47.4llama.cpp · Q4_K_M
Reproduce this run
toolllama.cpp b11985-6d194531e
modelLlama-3.1-Tulu-3-8B-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 "Llama-3.1-Tulu-3-8B-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.