NVIDIA-Nemotron-3.5-Super-VL-120B-A12B
7 records across 3 hardware.
| hardware | model | quant | backend | tok/s | pp tok/s | W | ttft s | ctx/tg | date | class | flags | source |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| M5 Ultra | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | 4bit | oMLX | 72.0 | 769.3 | – | – | 1024 | 2026-10-17 | community | – | source |
| M5 Ultra | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | 4bit | oMLX | 71.5 | 1758.0 | – | – | 4096 | 2026-10-17 | community | – | source |
| M5 Ultra | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | 4bit | oMLX | 59.5 | 1574.0 | – | – | 131072 | 2026-10-17 | community | – | source |
| M4 Max | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | 4bit | oMLX | 50.8 | 400.7 | – | – | 1024 | 2026-10-18 | community | – | source |
| M4 Max | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | 4bit | oMLX | 50.2 | 426.8 | – | – | 4096 | 2026-10-18 | community | – | source |
| M1 Ultra | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | 4bit | oMLX | 41.7 | 328.5 | – | – | 4096 | 2026-10-16 | community | – | source |
| M1 Ultra | NVIDIA-Nemotron-3.5-Super-VL-120B-A12B | 4bit | oMLX | 40.1 | 343.2 | – | – | 16384 | 2026-10-16 | community | – | source |