Benchmark result
swift-1.5-9b-q6_k on AMD Radeon RX 9050 / 9060 XT — 39.7 tok/s
Measured with llama.cpp manual on November 5, 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 3 runs of this model on AMD Radeon RX 9050 / 9060 XT · median 39.7 tok/s.
sort
- 139.7tok/s83.0llama.cpp · Q6_K
- 239.7tok/s65.2llama.cpp · Q6_Kthis run
- 339.6tok/s74.9llama.cpp · Q6_K
Reproduce this run
toolllama.cpp manual
modelswift-1.5-9b-q6_k (Q6_K)
context200,000
thinkingon
think limit16,384 tokens
samplingtemperature 1 · top_p 0.949999988079071 · top_k 20 · min_p 0 · presence_penalty 1.5 · repeat_penalty 1
clientv0.4.98+102
Set those in llama.cpp, then:
llm-benchmark benchmark --model "swift-1.5-9b-q6_k" --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.