# Deploy NVIDIA-Nemotron-3-Super-120B-A12B-FP8 on dedicated GPUs

NVIDIA-Nemotron-3-Super-120B-A12B-FP8 by NVIDIA, deployable on sference as a dedicated deployment: reserved GPU capacity, private endpoint, European GPUs, billed per GPU-hour on monthly commitments. Request it at https://sference.com/models/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8.

## Specs

- Hugging Face: https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8
- Category: Chat & reasoning
- Parameters: 124B (12B active)
- Architecture: MoE · 512 experts, 22 active
- Context: 256K tokens
- Published precision: FP8 + BF16
- Checkpoint size: 128 GB
- Minimum footprint: 2 GPUs (80 GB) or 2 GPUs (141 GB) to load the weights; more for long contexts and concurrency
- License: NVIDIA Nemotron Open Model License (Commercial use allowed) — https://www.nvidia.com/en-us/agreements/enterprise-software/nvidia-nemotron-open-model-license/

## FAQ

### Can I run NVIDIA-Nemotron-3-Super-120B-A12B-FP8 on a dedicated deployment in Europe?

Yes. sference deploys NVIDIA-Nemotron-3-Super-120B-A12B-FP8 on GPU capacity reserved for you, behind a private endpoint, with every request processed on European GPUs and logged in your audit trail. Tell us your expected volume and latency targets with the form on this page and we come back with a sizing and a quote.

### How many GPUs does NVIDIA-Nemotron-3-Super-120B-A12B-FP8 need?

The published checkpoint is about 128 GB (FP8 + BF16), so loading it takes at least 2 GPUs with 80 GB of memory each, or 2 GPUs with 141 GB each. Long contexts and high concurrency need more for the KV cache; quantization needs less. You don't pick hardware on sference: you choose a performance tier and we size the deployment.

### Is NVIDIA-Nemotron-3-Super-120B-A12B-FP8 licensed for commercial use?

NVIDIA-Nemotron-3-Super-120B-A12B-FP8 is published under the NVIDIA Nemotron Open Model License license. Commercial use allowed under NVIDIA's open model terms; keep the attribution notices.

### How is a dedicated deployment priced?

Per GPU-hour on a monthly commitment, sized to your throughput and latency targets.

Facts from Hugging Face, synced 2026-10-27. All deployable models: https://sference.com/dedicated.md