Nemotron 3 Super
Nemotron 3 Super is an open-weight language model from NVIDIA, released on March 11, 2026. It is a mixture-of-experts model with 120B total parameters, of which 12B are active for each token. NVIDIA lists a context window of up to 1M tokens. The model takes text input.
- Developer
- NVIDIA
- Released
- March 11, 2026
- Total parameters
- 120B
- Active parameters per token
- 12B
- Architecture
- MoE
- Context window
- 1M
- Input
- Text
- License
- NVIDIA Nemotron Open Model License
- Commercial use
- Yes
- Official weights, GB
- 247.2 GB
- Hugging Face repository
- huggingface.co
- Checked on
The weights are released under the NVIDIA Nemotron Open Model License, which allows commercial use.
The official weights on Hugging Face take about 247 GB in BF16. Running the model needs at least that much memory across GPUs and system RAM, plus room for the context cache; quantized versions need less.
- Total parameters, billions
- 120B
- Active parameters, billions
- 12B
- Context window, tokens
- 1M
- License type
- Custom license
- License conditions
- Commercial use is allowed. Redistributions must include the license and the attribution "Licensed by NVIDIA Corporation under the NVIDIA Nemotron Model License"; patent or copyright litigation ends the license, and export controls apply.
- Weights precision
- BF16
- Official quantized or alternative versions
- nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8; nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4
- License text
- nvidia.com
- Sources
Page URL Hugging Face model card https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16/raw/main/README.md License text (nvidia.com) https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-nemotron-open-model-license/ Hugging Face weights index https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16/raw/main/model.safetensors.index.json