Benchmark result
Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S-mtp on AMD Radeon RX 7800 XT — 48.9 tok/s
Measured with llama.cpp b10805-86b351fd6 on October 4, 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
1st fastest of 5 runs of this model on AMD Radeon RX 7800 XT · median 37.7 tok/s.
sort
- 148.9tok/s79.7llama.cppthis run
- 239.6tok/s87.4llama.cpp k5 v5 16384 budget
- 337.7tok/s88.1llama.cpp
- 437.2tok/s79.0llama.cpp
- 534.6tok/s83.1llama.cpp
Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S-mtp on other hardware
Reproduce this run
toolllama.cpp b10805-86b351fd6
modelSwift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S-mtp
context128,000
effortxhigh (model default)
think limitnone
samplingtemperature 1 · top_p 0.949999988079071 · top_k 20 · min_p 0 · presence_penalty 0 · repeat_penalty 1
clientv0.4.87+102
Set those in llama.cpp, then:
llm-benchmark benchmark --model "Swift-1.5-Qwen3.8-27B-GSQ-RCO-IQ3_S-mtp" --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.