Models /

h2o-danube3-500m-chat

31results
1devices
0.51Bparams

llama.cpp

31 tiny/test model result(s) (excluded from headlines)
HardwareQuantGen tok/sPrompt tok/sSourceContext testedBuild
RTX 4050 Laptop GPUQ4_K - Medium21910,670LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ2_K - Medium2119,347LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ2_K - Small2108,760LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ4_120610,482LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ5_020410,493LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ4_020410,661LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ3_K - Small2049,703LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ4_K - Small20310,347LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ5_K - Small20210,384LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ3_K - Medium20110,136LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ5_K - Medium20010,035LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ4_K - Medium19810,530LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ3_K - Large19610,154LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ5_119610,424LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ6_K1899,569LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUunknown, may not work17510,676LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUQ8_01739,982LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUF161369,907LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUall F3280.36,559LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUBF1647.85,258LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ1_S - 1.5625 bpw38.6150LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ2_XS - 2.3125 bpw34.9143LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ2_M - 2.7 bpw30.5135LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ1_M - 1.75 bpw30.494.5LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ3_S mix - 3.66 bpw30.3122LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ2_S - 2.5 bpw30.1131LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ2_XXS - 2.0625 bpw28.9138LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ4_XS - 4.25 bpw28.4329LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ3_XXS - 3.0625 bpw26.3120LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ4_NL - 4.5 bpw26.3259LocalScore ↗512+128llamafile 0.9.3Details →
RTX 4050 Laptop GPUIQ3_S - 3.4375 bpw24.994.5LocalScore ↗512+128llamafile 0.9.3Details →

Ollama

No measured Ollama results yet. Submit yours →

MLX

No measured MLX results yet. Submit yours →

Includes results from LocalScore, a Mozilla Builders project by cjpais, mirrored in the cn0303/local-llm-speed-benchmarks dataset. LocalScore runs use llamafile, which is built on llama.cpp.