Hardware /
RTX 2000 Ada
11measured results
173best gen tok/s*
16 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 | 173 | 6,151 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 168 | 6,085 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 3.2 1B Instruct | Q4_K - Medium | 165 | 6,025 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Llama 2 7B | Q4_0 | 50.7 | 2,250 | llama.cpp (CUDA) ↗ | 512+128 | llama.cpp CUDA @756cfea | Details → |
| Llama 2 7B | Q4_0 | 50.6 | 1,956 | llama.cpp (CUDA) ↗ | 512+128 | llama.cpp CUDA @756cfea | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 37.9 | 1,328 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 37.9 | 1,373 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Meta Llama 3.1 8B Instruct | Q4_K - Medium | 36.8 | 1,370 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| YandexGPT 5 Lite 8B Instruct | Q4_K - Medium | 35.9 | 1,271 | LocalScore ↗ | 512+128 | llamafile 0.9.3 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 21.0 | 733 | LocalScore ↗ | 512+128 | llamafile 0.9.2 | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 20.3 | 748 | 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.