Llama 2 7B on Apple M1
12.8best gen tok/s
38.3best prompt tok/s
640max context tested
2results
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 | 12.8 | 35.9 | llama.cpp (Vulkan) ↗ | 512+128 | — | llama.cpp Vulkan @2370665 — Asahi Linux |
| Q4_0 | 12.5 | 38.3 | llama.cpp (Vulkan) ↗ | 512+128 | — | llama.cpp Vulkan @2370665 — Asahi Linux |
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.