N
Nemotron 3 Super (free)OPEN
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.
Prices update automatically — checked daily against provider list prices.
- Free — open weights, no per-token cost.
- 3 of 10 NVIDIA models tracked here, from cheapest
Benchmarks & performance
Independent benchmark scores for Nemotron 3 Super (free), measured by Artificial Analysis. Higher is better. Measurement mode: Reasoning.
For the same budget, Nemotron 3 Super (free) delivers 12 times more capability. Open weights also mean you can self-host it and pay nothing per token.
About this model
Nemotron 3 Super (free) is an open-weight AI model by NVIDIA. You can download and self-host it for free; the prices below are hosted-API list prices, tracked daily, for when you prefer convenience over self-hosting.
Frequently asked questions
What is Nemotron 3 Super (free)?
Nemotron 3 Super (free) is an AI language model from NVIDIA. It is open-weight: you can download it and run it on your own hardware, for free. It scores 25.7 on the Artificial Analysis intelligence index.
Is Nemotron 3 Super (free) free?
The weights are free and open — you can self-host Nemotron 3 Super (free) and pay nothing per token. If you prefer a hosted API, list prices are Free per million input tokens and Free per million output tokens.
What is Nemotron 3 Super (free) good at?
Independent benchmarks from Artificial Analysis give it GPQA 80%, Humanity's Last Exam 20.8%, Long Context Reasoning 60.3%, SciCode 36%, IFBench 71.5%, τ²-Bench 67.8%, τ-Bench Banking 10.3%, Terminal-Bench 38.6%, Terminal-Bench Hard 28.8%. It is particularly used for code generation.
How fast is Nemotron 3 Super (free)?
It generates about 144.2 tokens per second, with a median 1.11s delay before the first token. Measured independently by Artificial Analysis.
Can I self-host Nemotron 3 Super (free)?
Yes. Nemotron 3 Super (free) has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.