--- title: "NVIDIA Nemotron family — distinct SKUs · Undominated.ai" canonical: https://undominated.ai/families/nemotron/ description: "Lightning, Nano, Super, Ultra and content-safety checkpoints do different jobs. The Nemotron-Flash research family is not evidence of a priced Flash endpoint in this catalogue." --- # NVIDIA Nemotron family — distinct SKUs · Undominated.ai > Lightning, Nano, Super, Ultra and content-safety checkpoints do different jobs. The Nemotron-Flash research family is not evidence of a priced Flash endpoint in this catalogue. NVIDIA / portfolio # NVIDIA Nemotron family Lightning, Nano, Super, Ultra and content-safety checkpoints do different jobs. The Nemotron-Flash research family is not evidence of a priced Flash endpoint in this catalogue. Family research reviewed 2026-09-22 · Official source notes below Standard models 5 Independently rated 0 / 5 Known token rates 5 / 5 Global balanced frontier 0 [Compare the members ↓](#members)[Read source notes](#source-notes)[All families](https://undominated.ai/families/) [Compare models](https://undominated.ai/compare/) Comparison context Choose up to three standard models in Compare. Requirements and prompt length apply in Compare. This page keeps its stated price basis and workload controls. ## Before you choose ### Research name versus priced row The Nemotron-Flash paper establishes a research family. It does not establish an available priced endpoint. NVIDIA also distinguishes model weights from NeMo tooling. [Nemotron-Flash: Towards Latency-Optimal Hybrid Small Language Models ↗](https://arxiv.org/abs/2511.18890) [NVIDIA Nemotron family guide ↗](https://www.nvidia.com/en-us/ai-data-science/foundation-models/nemotron/llm-info/) ## Compare exact catalogue rows USD per 1M tokens. In / out are listed base rates; expand billing conditions before planning a request. LMArena scores measure the general preference board, not success on your engineering task. Unknown stays unknown. ### Content safety | Model | LMArena | Input / output $/M | Context / max output | Capabilities & conditions | | --- | --- | --- | --- | --- | | [Nemotron 3.5 Content Safety](https://undominated.ai/models/nvidia__nemotron-3.5-content-safety/) [Check](https://undominated.ai/check/nvidia__nemotron-3.5-content-safety/) [Compare](https://undominated.ai/compare/?models=nvidia%2Fnemotron-3.5-content-safety) [Full evidence ↗](https://undominated.ai/models/nvidia__nemotron-3.5-content-safety/) nvidia/nemotron-3.5-content-safety | Unrated | $0.200 / $0.200 Listed base rates Billing details Cached input Unknown Separate reasoning Not separately listed Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence. | 131,072 Output 117,964 | text · image Tools no · reasoning yes Weights unknown Released 2026-06-04 | ### Reasoning and agentic checkpoints | Model | LMArena | Input / output $/M | Context / max output | Capabilities & conditions | | --- | --- | --- | --- | --- | | [Nemotron 3.5 Lightning](https://undominated.ai/models/nvidia__nemotron-3.5-lightning/) [Check](https://undominated.ai/check/nvidia__nemotron-3.5-lightning/) [Compare](https://undominated.ai/compare/?models=nvidia%2Fnemotron-3.5-lightning) [Full evidence ↗](https://undominated.ai/models/nvidia__nemotron-3.5-lightning/) nvidia/nemotron-3.5-lightning | Unrated | $0.060 / $0.160 Listed base rates Billing details Cached input $0.030 Separate reasoning Not separately listed Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence. | 262,144 Output 32,768 | text Tools yes · reasoning yes Weights unknown Released 2026-08-11 | | [Nemotron 3 Ultra](https://undominated.ai/models/nvidia__nemotron-3-ultra-550b-a55b/) [Check](https://undominated.ai/check/nvidia__nemotron-3-ultra-550b-a55b/) [Compare](https://undominated.ai/compare/?models=nvidia%2Fnemotron-3-ultra-550b-a55b) [Full evidence ↗](https://undominated.ai/models/nvidia__nemotron-3-ultra-550b-a55b/) nvidia/nemotron-3-ultra-550b-a55b | Unrated | $0.625 / $3.13 Listed base rates Billing details Cached input $0.188 Separate reasoning Not separately listed Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence. | 262,144 Output 16,384 | text Tools yes · reasoning yes Open weights Released 2026-06-04 | | [Nemotron 3 Super](https://undominated.ai/models/nvidia__nemotron-3-super-120b-a12b/) [Check](https://undominated.ai/check/nvidia__nemotron-3-super-120b-a12b/) [Compare](https://undominated.ai/compare/?models=nvidia%2Fnemotron-3-super-120b-a12b) [Full evidence ↗](https://undominated.ai/models/nvidia__nemotron-3-super-120b-a12b/) nvidia/nemotron-3-super-120b-a12b | Unrated | $0.085 / $0.400 Listed base rates Billing details Cached input Unknown Separate reasoning Not separately listed Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence. | 262,144 Output 235,929 | text Tools yes · reasoning yes Weights unknown Released 2026-03-11 | | [Nemotron 3 Nano 30B A3B](https://undominated.ai/models/nvidia__nemotron-3-nano-30b-a3b/) [Check](https://undominated.ai/check/nvidia__nemotron-3-nano-30b-a3b/) [Compare](https://undominated.ai/compare/?models=nvidia%2Fnemotron-3-nano-30b-a3b) [Full evidence ↗](https://undominated.ai/models/nvidia__nemotron-3-nano-30b-a3b/) nvidia/nemotron-3-nano-30b-a3b | Unrated | $0.050 / $0.200 Listed base rates Billing details Cached input $0.030 Separate reasoning Not separately listed Tool, media, cache-write and service-tier charges may add dimensions. Read the full model evidence. | 262,144 Output 235,929 | text Tools yes · reasoning yes Weights unknown Released 2025-12-14 | ## Free and batch variants These rows have different access or processing conditions. They are excluded from the standard-row counts above. A variant without a standard base remains a separately named variant. - [Nemotron 3.5 Lightning (free)](https://undominated.ai/models/nvidia__nemotron-3.5-lightning__free/) free · free in / free out per 1M - [Nemotron 3.5 Content Safety (free)](https://undominated.ai/models/nvidia__nemotron-3.5-content-safety__free/) free · free in / free out per 1M - [Nemotron 3 Ultra (free)](https://undominated.ai/models/nvidia__nemotron-3-ultra-550b-a55b__free/) free · free in / free out per 1M - [Nemotron 3 Nano Omni (free)](https://undominated.ai/models/nvidia__nemotron-3-nano-omni-30b-a3b-reasoning__free/) free · free in / free out per 1M - [Nemotron 3 Super (free)](https://undominated.ai/models/nvidia__nemotron-3-super-120b-a12b__free/) free · free in / free out per 1M ## Named, but absent from the accepted catalogue - [Nemotron Flash](https://undominated.ai/models/nvidia__nemotron-flash/) — No accepted catalogue entry for Nemotron Flash ## Sources and scope - [NVIDIA Nemotron family guide ↗](https://www.nvidia.com/en-us/ai-data-science/foundation-models/nemotron/llm-info/) Exact named models on the vendor platform; not a promise about every reseller Checked 2026-09-22 · official document opened and inspected - [Nemotron-Flash: Towards Latency-Optimal Hybrid Small Language Models ↗](https://arxiv.org/abs/2511.18890) Research model family, not an inference API listing Checked 2026-09-22 · primary paper abstract opened and inspected Source review does not refresh a price or transplant a benchmark score. [Read the method](https://undominated.ai/methodology/) and [correction policy](https://undominated.ai/corrections/). People searched nvidia nemotron flash, nemotron 3 super, nemotron 3 ultra. ## Questions this page answers Are the NVIDIA Nemotron family models interchangeable? No. These are exact catalogue rows with separate capability, pricing and rating evidence. Compare the checkpoint, serving endpoint, reasoning settings and required input types. A family name does not establish equivalent behaviour. How many NVIDIA rows are compared here? 5 standard rows, of which 0 have an accepted LMArena general score and 5 have known input and output token rates. 5 related free or batch variants are listed separately. What do the prices and frontier counts mean? Prices below are listed US dollars per million input or output tokens, before request-specific context tiers and other billing dimensions. The frontier count comes from the global balanced view. Neither a family price range nor a benchmark difference proves your task will be cheaper or more successful. How current are the official notes? The linked family research was reviewed 2026-09-22. That date is separate from each accepted catalogue observation and each vendor’s effective date. An official note applies to its named platform, not automatically to every reseller. ## Continue your investigation - [Compare exact models](/compare/) - [Inspect model evidence](/models/) - [Check lifecycle notices](/deprecations/) - [Estimate local capacity](/self-host/)