Titan V

7measured results

232best gen tok/s*

12 GBreported memory

*Headline excludes tiny/test models (under 0.6B params or TinyStories-style). Open in rig checker →

llama.cpp

Model	Quant	Gen tok/s	Prompt tok/s	Source	Context tested	Build
Llama 3.2 1B Instruct	Q4_K - Medium	232	7,853	LocalScore ↗	512+128	llamafile 0.9.2	Details →
Llama 2 7B	Q4_0	112	2,481	llama.cpp (CUDA) ↗	512+128	llama.cpp CUDA @e56abd2	Details →
Llama 2 7B	Q4_0	109	793	llama.cpp (Vulkan) ↗	512+128	llama.cpp Vulkan @e56abd2	Details →
Llama 2 7B	Q4_0	109	2,617	llama.cpp (CUDA) ↗	512+128	llama.cpp CUDA @e56abd2	Details →
Llama 2 7B	Q4_0	105	796	llama.cpp (Vulkan) ↗	512+128	llama.cpp Vulkan @e56abd2	Details →
Meta Llama 3.1 8B Instruct	Q4_K - Medium	73.3	1,823	LocalScore ↗	512+128	llamafile 0.9.2	Details →
Qwen2.5 14B Instruct	Q4_K - Medium	42.2	962	LocalScore ↗	512+128	llamafile 0.9.2	Details →

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.