{"id":"nvidia/nvidia-nemotron-3-ultra-550b-a55b-bf16","aliases":["hf:nvidia/nvidia-nemotron-3-ultra-550b-a55b-bf16","nvidia/nemotron-3-ultra-550b-a55b","nvidia/nemotron-3-ultra-550b-a55b-20260604","nvidia/nemotron-3-ultra-550b-a55b:batch","nvidia/nemotron-3-ultra-550b-a55b:free","openrouter:nvidia/nemotron-3-ultra-550b-a55b","openrouter:nvidia/nemotron-3-ultra-550b-a55b:batch","openrouter:nvidia/nemotron-3-ultra-550b-a55b:free"],"displayName":"NVIDIA: Nemotron 3 Ultra","author":"nvidia","createdAt":"2026-06-04T05:33:28.000Z","updatedAt":"2026-09-08T17:00:59.308Z","popularityRank":20,"metadata":{"modalities":["text"],"inputModalities":["text"],"outputModalities":["text"],"contextLength":262144,"maxOutputTokens":32768,"family":"nemotron","modelsDev":{"provider":"openrouter","sourceUrl":"https://models.dev/api.json","description":"Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy","lastUpdated":"2026-06-04","releaseDate":"2026-06-04","capabilities":{"reasoning":true,"tool_call":true,"attachment":false,"temperature":true,"open_weights":true,"structured_output":true}},"tokenizer":"Other","licenseStatus":"unknown","weightsAvailable":false,"hostedInferenceAllowed":false,"gated":false,"supportedParameters":["frequency_penalty","include_reasoning","logit_bias","max_tokens","min_p","presence_penalty","reasoning","reasoning_effort","repetition_penalty","response_format","seed","stop","structured_outputs","temperature","tool_choice","tools","top_k","top_p"],"providerNames":[],"designArenaCategories":["3d","asciiart","codecategories","dataviz","gamedev","svg","uicomponent","website"],"isModerated":false},"availability":"unknown","externalBest":null,"multivibeNetwork":null,"description":"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...","runtimes":[],"protocols":[],"harnesses":[]}