NVIDIA Nemotron · General · Nemotron 3
NVIDIA Nemotron 3 Super 120B-A12B is a recent General release in the NVIDIA Nemotron family, certified from its immutable official source revision.
NVIDIA Nemotron 3 Super 120B-A12B is an official self-hostable NVIDIA release in the NVIDIA Nemotron General track. It uses NemotronHForCausalLM, carries 120B / 12B MoE, supports a verified context window of 262,144 tokens, and exposes only source-certified checkpoints and runtime combinations. Its release, license and deployment facts are tied to the official repository revision rather than inferred from the family name.
NVIDIA Nemotron 3 Super 120B-A12B uses NemotronHForCausalLM with 120B / 12B MoE. The stored architecture profile records 88 layers, hidden width 4096, 32 attention heads and a 262,144 token context ceiling where the official configuration publishes those fields.
Only exact, source-certified variants attached to this release are shown.
nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8
Official · FP8 · safetensors
Runtime certified: Transformers TensorRT-LLM
| Precision / quantization | Format | Bits | Publisher trust | Deployment state |
|---|---|---|---|---|
| FP8 | safetensors | 8 | Official | Runtime certified |
Use only the listed transformers, tensorrt-llm runtime configuration with the exact checkpoint precision. The recommendation engine checks weight memory, KV cache, runtime overhead, safety headroom, GPU architecture and topology before ranking rentable and reference hardware.
Recommended best value
240 GB total VRAM · nvswitch
50 GB headroom · Available
USD 5,580 / monthLowest-cost viable
282 GB total VRAM · nvswitch
92 GB headroom · Limited capacity
USD 4,690 / monthHighest performance
282 GB total VRAM · nvswitch
92 GB headroom · Limited capacity
USD 4,690 / monthThe full 120B resident weight set must fit across the selected GPUs. Context and concurrency add KV-cache memory; MoE active parameters do not replace resident-weight memory.
NVIDIA Nemotron 3 Super 120B-A12B is published under NVIDIA Nemotron Open Model License. The release-scoped official license link and verification date are retained with the catalogue record.
Commercial Use Allowed With Conditions. Commercial use is conditional. Review the linked official terms before deployment.
NVIDIA Nemotron 3 Ultra 550B-A55B, NVIDIA Nemotron 3.5 Lightning 30B-A3B, NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4
Source revision: 680e0e5a3131d196ec831e0f1d320c7e0a5405faac89108daf9591babe14134c
The exact estimate depends on checkpoint precision, context and concurrency; the page recommendation uses the certified memory engine and compatible runtime configurations.
Certified upstream runtime evidence is available for transformers, tensorrt-llm.
The stored commercial classification is Commercial Use Allowed With Conditions. Consult the official NVIDIA Nemotron Open Model License terms for the final obligations.
Last verified: 2026-08-26
