Benchmark result
GRM-3.2-Sky on 2× Tesla V100-PCIE-32GB — 94.4 tok/s
Measured with llama.cpp manual on November 3, 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
Reproduce this run
toolllama.cpp manual
modelGRM-3.2-Sky
context32,768
think limit8,096 tokens
samplingtemperature 0.5 · top_p 0.8999999761581421 · 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 "GRM-3.2-Sky" --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.