NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. Built on a hybrid Mamba-Transformer Mixture-of-Experts architecture with multi-token prediction (MTP), it delivers over 50% higher token generation compared to leading open models. The model features a 1M token context window for long-term agent coherence, cross-document reasoning, and multi-step task planning. Latent MoE enables calling 4 experts for the inference cost of only one, improving intelligence and generalization. Multi-environment RL training across 10+ environments delivers leading accuracy on benchmarks including AIME 2025, TerminalBench, and SWE-Bench Verified. Fully open with weights, datasets, and recipes under the NVIDIA Open License, Nemotron 3 Super allows easy customization and secure deployment anywhere — from workstation to cloud.
Public pricing comes from the currently exposed Xenodia model catalog and already reflects the minimum visible channel pricing.
Use this model when you need a public Xenodia surface with discoverable parameters, stable routing metadata, and documented request modes instead of provider-specific guesswork.
Recommended workflow: query /v1/models, inspect this model's capabilities, then call the relevant endpoint with only documented fields.
curl -s https://api.xenodia.xyz/v1/modelsNo public parameter metadata has been published for this model yet.
This model does not currently advertise public reference input support.
nemotron-3-super-120b-a12b:free is exposed through Xenodia as a public model page with pricing, modalities, parameters, and documented request modes.
nemotron-3-super-120b-a12b:free does not currently advertise public task mode on Xenodia.