llama 1B
19 records across 2 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M4 | llama 1B | Q4_0 | llama.cpp (Metal) | 121.2 | 1334.54 | – | – | tg128 | 2024-05-20 | community | – | source |
| M4 | llama 1B | Q8_0 | llama.cpp (Metal) | 75.33 | 1333.82 | – | – | tg128 | 2024-05-20 | community | – | source |
| M4 | llama 1B | F16 | llama.cpp (Metal) | 46.21 | 1375.87 | – | – | tg128 | 2024-05-20 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 12.53 | – | – | – | – | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 9.02 | – | – | – | – | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 5.39 | – | – | – | 1024 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 4.61 | – | – | – | 1024 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 3.79 | – | – | – | 2048 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 3.34 | – | – | – | 2048 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 2.32 | – | – | – | 4096 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | 2.18 | – | – | – | 4096 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 25.81 | – | – | – | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 249.62 | – | – | – | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 22.85 | – | – | 1024 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 107.69 | – | – | 1024 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 20.41 | – | – | 2048 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 68.02 | – | – | 2048 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 16.89 | – | – | 4096 | 2026-10-02 | community | – | source |
| IBM Z Neural Network Processing Assist (NNPA) | llama 1B | Q8_0 | llama.cpp (zDNN) | – | 40.41 | – | – | 4096 | 2026-10-02 | community | – | source |