NVIDIA Nemotron 3 Super 120B A12B (Reasoning)
Open weightsOpen weightsText120B total / 12B active (MoE)Reasoning
Nemotron family model by NVIDIA — reasoning.
About the Nemotron family
NVIDIA's open Nemotron family — Llama-derived and original models (Nano ~4–12B, Super ~49–120B, Ultra up to 550B MoE) heavily optimized via pruning, distillation and reasoning post-training, plus vision (VL) and omni variants. Designed to showcase NVIDIA hardware and ship with open training data recipes.
All Nemotron modelsHow this one is configured
This configuration runs in reasoning mode (thinks step-by-step before answering); mixture-of-experts with ~120B total / ~12B active parameters per token.
Who makes it
NVIDIA — United States. Not listed as a provider in this directory.
More from Nemotron
- Llama 3.1 Nemotron Instruct 70B
- Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)
- Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)
- Llama 3.3 Nemotron Super 49B v1 (Non-reasoning)
- Llama 3.3 Nemotron Super 49B v1 (Reasoning)
- Llama Nemotron Super 49B v1.5 (Non-reasoning)
- Llama Nemotron Super 49B v1.5 (Reasoning)
- Nemotron 3 Nano Omni 30B A3B Reasoning
- Nemotron 3 Ultra 550B A55B (Reasoning)
- Nemotron Cascade 2 30B A3B
- NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)
- NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)
- NVIDIA Nemotron 3 Nano 4B
- NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)
- NVIDIA Nemotron Nano 12B v2 VL (Reasoning)
- NVIDIA Nemotron Nano 9B V2 (Non-reasoning)
- NVIDIA Nemotron Nano 9B V2 (Reasoning)