Nemotron 3 Ultra 550B A55B: pricing, context and performance

Nemotron 3 Ultra 550B A55B at a glance

Nemotron 3 Ultra 550B A55B is a language model from NVIDIA, released on June 4, 2026. That makes it 4 months old as of October 22, 2026. It is an open-weight model: you can host it yourself on infrastructure you control, and 25 API providers listed on models.dev also offer it. It has a reasoning mode: it can "think" before answering, which helps with multi-step problems but uses more tokens, and therefore more money.

We file it under open-weight models (68 models in this category on the site).

What it costs

Nemotron 3 Ultra 550B A55B costs $0.50 per million input tokens and $2.4 per million output tokens (median of 25 providers, checked on October 22, 2026).

In practice, an agent task that reads 20,000 tokens and writes 2,000 costs about $0.015, or $14.8 for 1,000 such tasks.

That is 2.4 times the median output price of open-weight models ($1).

To estimate your own budget, use the agent cost calculator.

Benchmark results

As of October 22, 2026, Epoch AI publishes no score for Nemotron 3 Ultra 550B A55B on our basket's benchmarks. To place it, rely on its specifications and price, then test it on your own cases.

What it can do

Its context window reaches 1,000,000 tokens, roughly 1,500 pages of text at 500 words per page: enough to load a whole book or a mid-sized codebase. A single answer can run to 65,536 tokens.

It accepts as input: text. Its knowledge stops at September 2025: beyond that, it needs web search or your documents.

It can call tools (functions, APIs) and return structured JSON output, which makes it a possible foundation for an AI agent.

The weights are published on Hugging Face (nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16): 560.5 billion parameters, other license and 1,058,710 downloads over the last 30 days. See also our list of open-source models.

Who is it for?

Given its category, Nemotron 3 Ultra 550B A55B is mainly aimed at projects where you want control over your data or hosting, or want to adapt the model to your field.

Its very large context window makes it worth considering for analysing large sets of documents in a single request.

Before adopting it, test it on your own examples: benchmarks measure standard tasks, not your job.

What could replace it?

In the same category, the models closest to Nemotron 3 Ultra 550B A55B are:

• Kimi K2 Thinking (Moonshot AI, index 69.6)
• Kimi K2 0905 (Moonshot AI, $2.5 per million output tokens)
• Hy4 preview (Tencent (Hunyuan), $2.5 per million output tokens)
• DeepSeek R1 0528 (DeepSeek, $2.21 per million output tokens)