Hardware /
Tesla T4
3measured results
46.4best gen tok/s*
15 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 2 7B | Q4_0 | 46.4 | 1,219 | llama.cpp (CUDA) ↗ | 512+128 | llama.cpp CUDA @d32e03f | Details → |
| Llama 2 7B | Q4_0 | 44.0 | 1,310 | llama.cpp (CUDA) ↗ | 512+128 | llama.cpp CUDA @d32e03f | Details → |
| Qwen2.5 14B Instruct | Q4_K - Medium | 13.3 | 450 | 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.