Llama-3.2-3B
7 records across 2 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M4 Max (128 GB) | Llama-3.2-3B | Q8_0 | llama.cpp | 34.0 | – | – | – | 4096 | 2026-10-17 | sourced | – | source |
| IQ-9075 | Llama-3.2-3B | Q2_K | llama.cpp (Hexagon) | 27.33 | – | – | – | – | 2026-10 | community | – | source |
| IQ-9075 | Llama-3.2-3B | Q2_K | llama.cpp (Hexagon) | 26.18 | – | – | – | – | 2026-10 | community | – | source |
| IQ-9075 | Llama-3.2-3B | Q2_K | llama.cpp (Hexagon) | 24.92 | – | – | – | – | 2026-10 | community | – | source |
| IQ-9075 | Llama-3.2-3B | Q2_K | llama.cpp (Hexagon) | 24.47 | – | – | – | – | 2026-10 | community | – | source |
| M4 Max (128 GB) | Llama-3.2-3B | Q8_0 | llama.cpp | 12.8 | – | – | – | 32768 | 2026-10-17 | sourced | – | source |
| M4 Max (128 GB) | Llama-3.2-3B | Q8_0 | llama.cpp | 5.9 | – | – | – | 102400 | 2026-10-17 | sourced | – | source |