model: qwen3-14b

4 runs · grouped below by site-grade

See also: on pavilion+predator (cross-machine) family: gemma/qwen3 task: chat all benchmarks

Flagship   Flagship (1)

TitleDateHardwareEngineModelsCallsHeadline
Flagship Predator Qwen rerun matrix 2026-05-04 predator llamacpp qwen3.5:9b, qwen3:14b 36 14.2 tok/s

Standard   Standard (1)

TitleDateHardwareEngineModelsCallsHeadline
Standard Pascal Vulkan-vs-CUDA Cross-Machine Benchmark — 2026-05-06 2026-05-06 pavilion+predator (cross-machine) llamacpp gemma4-e4b, gemma4-e2b, qwen35-0.8b, qwen35-2b, qwen35-4b, qwen35-9b, qwen3-14b, qwen3-30b-a3b-iq2m-moe:cuda:ngl999:n-cpu-moe35, qwen3-30b-a3b-iq2m-moe:vulkan:ngl999:n-cpu-moe35 26 —

Archive   Archive-only (2)

TitleDateHardwareEngineModelsCallsHeadline
Archive pavilion-weeyuga-v3 — qwen2.5/qwen2.5-coder/qwen3/qwen3.5 on pavilion 2026-04-29 pavilion weeyuga qwen3.5:4b, qwen3.5:35b-a3b-uncensored, qwen3.5:35b-a3b-iq2s, qwen3.5:9b-q6k, qwen3.5:9b, qwen3.5:2b, qwen3.5:0.8b, qwen2.5-coder:14b, qwen2.5-coder:3b, qwen3:14b, qwen3:8b, qwen3:4b, qwen2.5:3b, qwen2.5-coder:1.5b, qwen2.5-coder:0.5b 96 (12 err) —
Archive pavilion-weeyuga-v1 — qwen2.5/qwen2.5-coder/qwen3/qwen3.5 on pavilion 2026-04-28 pavilion weeyuga qwen3.5:4b, qwen3.5:35b-a3b-uncensored, qwen3.5:35b-a3b-iq2s, qwen3.5:9b-q6k, qwen3.5:9b, qwen3.5:2b, qwen3.5:0.8b, qwen2.5-coder:14b, qwen2.5-coder:3b, qwen3:14b, qwen3:8b, qwen3:4b, qwen2.5:3b, qwen2.5-coder:1.5b, qwen2.5-coder:0.5b 17 (3 err) —