tokenstat
tokenstat

Nemotron 3 Ultra 550B A55B

nvidia/nemotron-3-ultra-550b-a55b · rates as of · Benchmarks · Plans · Calculator

NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

Cheapest input
$0
$/MTok · models.dev
Cheapest output
$0
$/MTok
Context
1M
tokens
Offers
6
2 sources

Offers

Same model, different listings. First-party, router, and aggregator rows stay separate so you can see the spread instead of a single blended number. Showing free/$0 offers. Hide them.

SourceProviderInputOutputCache readContext
models.devcheapestnemotron-3-ultra-550b-a55bKenari$0$0$01M
models.devnvidia/nemotron-3-ultra-550b-a55bKilo Gateway$0.5$2.20$0.1512K
models.devnvidia/nemotron-3-ultra-550b-a55bNVIDIA$0.5$2.50$0.151M
models.devnvidia/nemotron-3-ultra-550b-a55bVercel AI Gateway$0.6$2.40$0.121M
OpenRouternvidia/nemotron-3-ultra-550b-a55bOpenRouter$0.6$3.60$0.2512K
models.devnvidia/nemotron-3-ultra-550b-a55bTogether AI$0.6$3.60$0.2512K

Capabilities

  • Reasoning
  • Tool calling
  • Structured output
  • Open weights
  • In: text
  • Out: text
Released: 2026-06-04Family: nemotron

Scores

Sparse coverage on purpose. Missing a score means we have not linked one yet, not that the model is unranked. Full boards on /benchmarks.

Sources: Arena AI, Artificial Analysis. See /benchmarks.

See what you actually spent

tokenstat prices your local sessions at list rates like these, with the cache split kept honest. Free, open source, runs on your machine.

$ curl -fsSL https://tokenstat.ai/install.sh | bash