NVIDIA Nemotron 3.5 Super VL: free downloadable NVFP4 model weights
- No end date
- on huggingface.co
Opens the official offer page.
The offer
Download NVIDIA Nemotron 3.5 Super VL 120B-A12B NVFP4 for self-hosted reasoning, coding, tool use and visual understanding. The OpenMDW-1.1 licence permits free commercial and non-commercial use globally, including India; compatible hardware or rented compute, storage and hosted services are separate costs. This is a free model download, not free hosted API inference or an included AI subscription.
Who can claim it
- Global deployment is documented; no India-specific exclusion found.
- Commercial and non-commercial use permitted under OpenMDW-1.1.
- Supply compatible hardware and sufficient storage; documented NVFP4 configurations include DGX Spark, H200, two H100 GPUs or GB200.
How to claim it
- Open the official NVIDIA Hugging Face model repository.
- Read the model card and linked OpenMDW-1.1 agreement.
- Download the model files using Files and versions or the serving instructions in the card.
- Follow the deployment recipe matching your supported hardware and runtime.
Good to know
No download-offer expiry is stated. Continued use is subject to OpenMDW-1.1; no hosted subscription is included.
Free model weights, not free hosted inference. The NVFP4 repository is approximately 75.1 GB. Hardware or cloud compute, electricity and storage are separate. Hosted API trials and enterprise services have separate terms. Redistribution requires the licence and applicable notices; third-party rights remain your responsibility. Licence rights can terminate under its litigation clause.
Source: the official offer page, huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4
Did it work for you?
Your vote is how a dead offer comes down: three or more "Didn't" in two days, outnumbering the "Worked" votes, and it is taken off the site.
Some links pay us a commission. It costs you nothing extra and never decides what is listed. Full disclosure.