Model by NVIDIA

Nemotron 3 Ultra

Open-weights model from NVIDIA. 262,144-token context. Listed at $0.6 in and $2.4 out per 1M tokens on OpenRouter (2026-10-31).

Checked 2026-10-31

Loading poll...

Nemotron 3 Ultra is NVIDIA's open-weights model, released 2026-06-04, and its listing name shows 550 billion total parameters with 55 billion active. OpenRouter lists $0.6 per million input tokens and $2.4 output, with a 262,144-token context and text-only input and output. Its licence in our data is open-weights, so check NVIDIA's terms before using it commercially. We would suggest it for teams that want a very large open model from a hardware vendor and have the capacity to host it, or want to use it through a gateway. The limit is size: a 550B model needs serious GPU memory to self-host, so most readers will use a hosted route, and the context is shorter than that of Nemotron 3 Super. We have not tested it ourselves. Prices and specifications come from the OpenRouter listing we checked on 31 October 2026, and we have not tested the model ourselves.

Mentioned here

Questions

What is the context window of Nemotron 3 Ultra?

262,144 tokens, with up to 182,520 tokens of output.

How much does Nemotron 3 Ultra cost?

$0.6 per million input tokens and $2.4 per million output tokens, checked 31 October 2026.

Can I run Nemotron 3 Ultra on my own hardware?

Yes, the weights are published under the Open-weights model licence. They are on huggingface.co.

Who makes Nemotron 3 Ultra?

NVIDIA makes it, and released it on 4 June 2026.

Loading poll...

Reviews

Loading reviews…

Write a review

Your rating

Up to 1000 characters. Stars alone are enough.

Part of the Wbcom Designs family

The all-in-one WordPress community stack

Also ours: wbcomdesigns.comvapvarun.combrndle.com