Llama 2 7B on RTX 4090
190best gen tok/s
14,771best 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 | 190 | 10,830 | llama.cpp (Vulkan) ↗ | 512+128 | — | llama.cpp Vulkan @4ae88d0 — coopmat2 |
| Q4_0 | 189 | 14,771 | llama.cpp (CUDA) ↗ | 512+128 | — | llama.cpp CUDA @2241453 |
| Q4_0 | 188 | 9,452 | llama.cpp (Vulkan) ↗ | 512+128 | — | llama.cpp Vulkan @4ae88d0 — coopmat2 |
| Q4_0 | 186 | 11,993 | llama.cpp (CUDA) ↗ | 512+128 | — | llama.cpp CUDA @2241453 |
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.