Nanbeige4.2-3B
14 records across 5 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M3 Ultra | Nanbeige4.2-3B | oQ4e | oMLX | 54.3 | 1127.0 | – | – | 1024 | 2026-10-08 | community | – | source |
| M3 Ultra | Nanbeige4.2-3B | oQ4e | oMLX | 46.6 | 1500.0 | – | – | 4096 | 2026-10-08 | community | – | source |
| M3 Ultra | Nanbeige4.2-3B | oQ4e | oMLX | 35.2 | 746.3 | – | – | 32768 | 2026-10-08 | community | – | source |
| M3 Ultra | Nanbeige4.2-3B | unknown | oMLX | 25.7 | 1515.0 | – | – | 4096 | 2026-10-08 | community | – | source |
| M3 Ultra | Nanbeige4.2-3B | unknown | oMLX | 25.2 | 1586.0 | – | – | 1024 | 2026-10-08 | community | – | source |
| M3 Pro | Nanbeige4.2-3B | OptiQ 4bit | oMLX | 24.3 | 387.2 | – | – | 1024 | 2026-10-09 | community | – | source |
| M3 Pro | Nanbeige4.2-3B | OptiQ 4bit | oMLX | 21.8 | 352.8 | – | – | 4096 | 2026-10-09 | community | – | source |
| M4 | Nanbeige4.2-3B | oQ4e | oMLX | 21.0 | 245.4 | – | – | 1024 | 2026-10-07 | community | – | source |
| M3 Ultra | Nanbeige4.2-3B | unknown | oMLX | 19.7 | 752.1 | – | – | 32768 | 2026-10-08 | community | – | source |
| M4 | Nanbeige4.2-3B | oQ4e | oMLX | 17.1 | 220.9 | – | – | 4096 | 2026-10-07 | community | – | source |
| M1 Pro | Nanbeige4.2-3B | 8bit | oMLX | 16.9 | 190.7 | – | – | 1024 | 2026-10-11 | community | – | source |
| AMD Custom GPU 0932 | Nanbeige4.2-3B 4.17B | Q4_K - Medium | llama.cpp (Vulkan) | 16.36 | 126.28 | – | – | tg128 | 2026-08-20 | community | – | source |
| AMD Custom GPU 0932 | Nanbeige4.2-3B 4.17B | Q4_K - Medium | llama.cpp (Vulkan) | 15.27 | 123.81 | – | – | tg128 | 2026-08-20 | community | – | source |
| M1 Pro | Nanbeige4.2-3B | 8bit | oMLX | 14.4 | 179.7 | – | – | 4096 | 2026-10-11 | community | – | source |