NVIDIA
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...
Running this yourself: likely needs a high-memory cloud gpu.
45.9
Quality Score
---
Arena ELO
120B
Parameters
1M
Context
Use this section to answer one simple question first: how much outside evidence do we have that this model performs well? Structured benchmark scores appear first, then official provider evidence, then live arena signal.
This model has normalized benchmark rows, so scores here are directly comparable across benchmark sources.
Sign in to join the discussion
0
Downloads
0
Likes
Mar 2026
Released
These are recent benchmark or leaderboard claims from official provider sources. They are useful for freshness and context, but they are not treated the same as normalized independent benchmark rows.
nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8 · Hugging Face
NVIDIA published benchmark or leaderboard evidence for Nemotron 3 Super, NVIDIA-Nemotron-3-Super-120B-A12B-FP8, NVIDIA Nemotron 3 Super 120B A12B FP8.
View sourceModels | Try NVIDIA NIM APIs
Nemotron Retriever + 4 Agentic Retrieval Code Retrieval Text-to-Embedding Retrieval Augmented Generation 13d Last updated on July 16, 2026 Thinkingmachines Downloadable Free Endpoint inkling Inkling is a multimodal (text + image) reasoning model from Thinking Machines — a Mamba-hybrid, 256-expert Mixture-of-Experts architecture with tool use and switchable reasoning. text-to-text + 3 reasoning image-to-text multimodal 14d Last updated on July 16, 2026 Poolside Free Endpoint l
View source