Benchmark result
unsloth/Qwen3.8-27B-GGUF:Q4_K_M on NVIDIA GeForce RTX 4090 — 97.8 tok/s
Measured with llama.cpp b10660-6c84c7d5d on August 29, 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
3rd fastest of 6 runs of this model on NVIDIA GeForce RTX 4090 · median 97.6 tok/s.
sort
- 199.3tok/s83.8llama.cpp · Q4_K_M
- 299.0tok/s85.3llama.cpp · Q4_K_M
- 397.8tok/s82.6llama.cpp · Q4_K_Mthis run
- 497.3tok/s87.5llama.cpp · Q4_K_M
- 590.3tok/s85.3llama.cpp · Q4_K_M
- 688.0tok/s86.2llama.cpp · Q4_K_M
Reproduce this run
toolllama.cpp b10660-6c84c7d5d
modelunsloth/Qwen3.8-27B-GGUF:Q4_K_M (Q4_K_M)
context131,072
clientv0.4.46+97
Set those in llama.cpp, then:
llm-benchmark benchmark --model "unsloth/Qwen3.8-27B-GGUF: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.