Hardware /
RTX 3080
40measured results
330best gen tok/s*
10 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 | 330 | 12,190 | LocalScore ↗ | 4,352 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 322 | 12,425 | LocalScore ↗ | 4,352 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 315 | 12,562 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 308 | 11,982 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 235 | 11,682 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 233 | 11,173 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 228 | 11,716 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 226 | 11,437 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 209 | 11,348 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 205 | 11,930 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 198 | 11,534 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 185 | 11,628 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 162 | 10,791 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 161 | 10,794 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 2 7B | Q4_0 | 146 | 4,914 | llama.cpp (Vulkan) ↗ | 512+128 | llama.cpp Vulkan @7c7d6ce — coopmat2 | Details → |
| Llama 2 7B | Q4_0 | 140 | 5,570 | llama.cpp (CUDA) ↗ | 512+128 | llama.cpp CUDA @9c35706 | Details → |
| Llama 2 7B | Q4_0 | 140 | 5,014 | llama.cpp (CUDA) ↗ | 512+128 | llama.cpp CUDA @9c35706 | Details → |
| Llama 2 7B | Q4_0 | 139 | 4,287 | llama.cpp (Vulkan) ↗ | 512+128 | llama.cpp Vulkan @7c7d6ce — coopmat2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 114 | 10,350 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 93.3 | 2,950 | LocalScore ↗ | 4,352 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 92.5 | 3,140 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 92.4 | 2,796 | LocalScore ↗ | 4,352 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 92.1 | 3,156 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 89.7 | 2,880 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 80.3 | 3,123 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 79.0 | 3,125 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Llama 3 8b Bnb 4bit | Q4_K - Medium | 78.3 | 3,204 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 76.6 | 3,028 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 76.4 | 3,128 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 75.7 | 2,792 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 75.5 | 2,891 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 74.8 | 3,079 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 68.2 | 2,836 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 52.1 | 2,963 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 52.0 | 1,595 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 46.6 | 1,741 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 41.0 | 1,307 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 31.1 | 874 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 30.7 | 1,169 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 14.4 | 257 | LocalScore ↗ | 4,352 | llamafile 0.9.3 | 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.