Llama 2 7B on RTX 2080 Ti
109best gen tok/s
3,108best prompt tok/s
640max context tested
4results
Measured · all numbers below come from the linked sources
llama.cpp
| Quant | Gen tok/s | Prompt tok/s | Source | Context tested | TTFT | Build |
|---|---|---|---|---|---|---|
| Q4_0 | 109 | 3,108 | llama.cpp (CUDA) ↗ | 512+128 | — | llama.cpp CUDA @9c35706 |
| Q4_0 | 108 | 2,891 | llama.cpp (CUDA) ↗ | 512+128 | — | llama.cpp CUDA @9c35706 |
| Q4_0 | 101 | 1,936 | llama.cpp (Vulkan) ↗ | 512+128 | — | llama.cpp Vulkan @N/A |
| Q4_0 | 97.6 | 1,888 | llama.cpp (Vulkan) ↗ | 512+128 | — | llama.cpp Vulkan @N/A |
Ollama
No measured Ollama results yet. Submit yours →
MLX
No measured MLX results yet. Submit yours →
"Context tested" is the largest prompt+generation length actually benchmarked (tokens), not a memory-fit estimate.