llm.ing

Nemotron 3 Ultra 550B A55B

Nvidiaopen weightsOpenRouter ↗Hugging Face ↗

Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy

Benchmarks

BenchmarkCategoryScoreVariantProvenanceSourceObserved
AA Agentic Indexagentic20.1mirroropenrouterSep 20, 2026
AA Coding Indexcoding49.3mirroropenrouterJul 23, 2026
AA Intelligence Indexintelligence22.9mirroropenrouterSep 20, 2026
Design Arena — 3dpreference1162mirroropenrouterSep 21, 2026
Design Arena — asciiartpreference1100mirroropenrouterSep 21, 2026
Design Arena — codecategoriespreference1156mirroropenrouterSep 21, 2026
Design Arena — datavizpreference1161mirroropenrouterSep 21, 2026
Design Arena — gamedevpreference1156mirroropenrouterSep 20, 2026
Design Arena — svgpreference1100mirroropenrouterSep 20, 2026
Design Arena — uicomponentpreference1154mirroropenrouterSep 20, 2026
Design Arena — websitepreference1149mirroropenrouterSep 20, 2026

Includes index data from Artificial Analysis.

Providers

Live operational stats per hosting endpoint, via OpenRouter.

ProviderQuantContext$/1M in · outUptime 24hLatencyThroughput
BaseTenfp4203K$0.6 · $2.499.97%
DeepInfrafp4262K$0.5 · $2.2
Venicefp8256K$0.625 · $3.1399.96%