NVIDIA-Nemotron-3-Super-120B-A12B-BF16
by nvidia
124B params · text-generation · 387 likes · 1.0M downloads
NVIDIA-Nemotron-3-Super-120B-A12B-BF16 is a 124B parameter model. At Q4 quantization it requires 62GB of VRAM. It runs comfortably on H100 SXM5 80 GB (39 tok/s), H200 SXM 141 GB (56 tok/s).
Inference providers
| Provider | $/1M in | $/1M out | Throughput |
|---|---|---|---|
| Featherless |
GPU compatibility
| GPU | VRAM | Q4 Decode | Verdict |
|---|---|---|---|
| M4 Max 128GB | 128GB | 3 tok/s | tight |
| A100 PCIe 80 GB | 80GB | 19 tok/s | tight |
| H100 SXM5 80 GB | 80GB | 39 tok/s | comfortable |
| H200 SXM 141 GB | 141GB | 56 tok/s | comfortable |
| M2 Ultra 192GB | 192GB | 5 tok/s | tight |