{
  "key": "nemotron-3-ultra-550b-a55b",
  "name": "Nemotron 3 Ultra 550B A55B",
  "description": "Largest Nemotron 3 model for maximum open-weight reasoning and agent accuracy",
  "lab": "nvidia",
  "family": "nemotron",
  "release_date": "2026-06-04",
  "knowledge": null,
  "open_weights": true,
  "has_free_offering": true,
  "published": true,
  "modalities": {
    "input": [
      "text"
    ],
    "output": [
      "text"
    ]
  },
  "capabilities": {
    "reasoning": true,
    "tool_call": true,
    "structured_output": true,
    "attachment": false,
    "reasoning_options": [
      {
        "type": "toggle"
      }
    ]
  },
  "limit": {
    "context": 1048576,
    "output": 1048576
  },
  "offerings": [
    {
      "provider": "baseten",
      "provider_model_id": "nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B",
      "name": "Nemotron Ultra",
      "variants": [
        {
          "org": "nvidia"
        }
      ],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.6,
        "output": 2.4,
        "cache_read": 0.12,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 202800,
        "output": 202800,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "bothub",
      "provider_model_id": "nemotron-3-ultra-550b-a55b:free",
      "name": "Nemotron 3 Ultra (free)",
      "variants": [
        "free"
      ],
      "free": true,
      "priced": false,
      "cost": {
        "input": 0,
        "output": 0,
        "cache_read": null,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 128000,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "edenai",
      "provider_model_id": "deepinfra/nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B (Deep Infra)",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.5,
        "output": 2.2,
        "cache_read": 0.1,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 262144,
        "output": 128000,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "edenai",
      "provider_model_id": "nebius/nvidia/Nemotron-3-Ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B (Nebius)",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 1,
        "output": 3,
        "cache_read": 1,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1048576,
        "output": 128000,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "kenari",
      "provider_model_id": "nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [],
      "free": false,
      "priced": false,
      "cost": {
        "input": 0,
        "output": 0,
        "cache_read": null,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 128000,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "kilo",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.5,
        "output": 2.2,
        "cache_read": 0.1,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 256000,
        "output": 32768,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "kilo",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b:free",
      "name": "NVIDIA: Nemotron 3 Ultra (free)",
      "variants": [
        "free"
      ],
      "free": true,
      "priced": false,
      "cost": {
        "input": 0,
        "output": 0,
        "cache_read": null,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 65536,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "nano-gpt",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b",
      "name": "Nvidia Nemotron 3 Ultra 550B",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.5,
        "output": 2.5,
        "cache_read": 0.25,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 65536,
        "input": 1000000
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "nano-gpt",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b:thinking",
      "name": "Nvidia Nemotron 3 Ultra 550B Thinking",
      "variants": [
        "thinking"
      ],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.5,
        "output": 2.5,
        "cache_read": 0.25,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 65536,
        "input": 1000000
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "nebius",
      "provider_model_id": "nvidia/Nemotron-3-Ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 1,
        "output": 3,
        "cache_read": 1,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1048576,
        "output": 1048576,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "nvidia",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.5,
        "output": 2.5,
        "cache_read": 0.15,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 65536,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "openrouter",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.625,
        "output": 3.125,
        "cache_read": 0.1875,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 262144,
        "output": 32768,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "openrouter",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b:free",
      "name": "Nemotron 3 Ultra (free)",
      "variants": [
        "free"
      ],
      "free": true,
      "priced": false,
      "cost": {
        "input": 0,
        "output": 0,
        "cache_read": null,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 65536,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "requesty",
      "provider_model_id": "nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [],
      "free": false,
      "priced": false,
      "cost": {
        "input": 0,
        "output": 0,
        "cache_read": null,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1048576,
        "output": 65536,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "togetherai",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.6,
        "output": 3.6,
        "cache_read": 0.2,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 512300,
        "output": 512300,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "unorouter",
      "provider_model_id": "nemotron-3-ultra-550b-a55b:free",
      "name": "Nemotron 3 Ultra 550B A55B",
      "variants": [
        "free"
      ],
      "free": true,
      "priced": false,
      "cost": {
        "input": 0,
        "output": 0,
        "cache_read": null,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 128000,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "venice",
      "provider_model_id": "nvidia-nemotron-3-ultra-550b-a55b",
      "name": "NVIDIA Nemotron 3 Ultra",
      "variants": [
        {
          "org": "nvidia"
        }
      ],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.625,
        "output": 3.125,
        "cache_read": 0.1875,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 256000,
        "output": 32768,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-11"
    },
    {
      "provider": "vercel",
      "provider_model_id": "nvidia/nemotron-3-ultra-550b-a55b",
      "name": "Nemotron 3 Ultra",
      "variants": [],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.6,
        "output": 2.4,
        "cache_read": 0.12,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 1000000,
        "output": 65000,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    },
    {
      "provider": "wandb",
      "provider_model_id": "nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B",
      "name": "Nemotron 3 Ultra",
      "variants": [
        {
          "org": "nvidia"
        }
      ],
      "free": false,
      "priced": true,
      "cost": {
        "input": 0.5,
        "output": 2.15,
        "cache_read": 0.1,
        "cache_write": null,
        "reasoning": null,
        "context_over_200k": null,
        "tiers": null
      },
      "limit": {
        "context": 262144,
        "output": 262144,
        "input": null
      },
      "status": null,
      "last_updated": "2026-06-04"
    }
  ]
}