Nemotron 3 Ultra 550B A55B

Nvidia llm

Flagship of Nvidia's Nemotron 3 family, a hybrid Mamba, mixture-of-experts and attention model for frontier reasoning and long-horizon agentic work.

overall ⌄

47.4

42 of 93 ranked

price · $/M tokens

self-hosted

open weights, no first-party API

to run it yourself ⌄

330 GB

minimum · 1.3 TB recommended

Benchmarks

What its maker published, and what anyone else measured. Every figure links the document it came from.

benchmarkpublishedmeasured
MMLU-Pro86.8
SWE-bench70.7
HLE26.7
BrowseComp44.4

Lineage

No recorded lineage — root or standalone model.

view in full graph →

Metadata

orgNvidia
released2026-06-04
params550B-A55B
licenseopenmdw-1.1
hf idnvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16