nvidia/nemotron-3-nano-omni/vision

Vision reasoning variant of NVIDIA's Nemotron 3 Nano Omni. 30B A3B hybrid Transformer-Mamba MoE - accepts an image plus a prompt and returns text.

Learn more about Nemotron

Inference

Commercial use

Schema

LLMs

Playground API Examples

Input

Prompt*

Image URL*

Hint: Drag and drop image files from your computer, images from web pages, paste from clipboard (Ctrl/Cmd+V), or provide a URL. Accepted file types: jpg, jpeg, png, webp, gif, avif

Additional Settings

Customize your input with more control.

Result

Idle

{
  "output": "This is a highly stylized, close-up portrait of a person featuring elaborate and surreal body art.\n\n**Face and Makeup:**\nThe person's face is the central focus, covered in intricate, colorful patterns that resemble a mix of tribal art, paisley, and organic textures.\n*   **Colors:** The",
  "finish_reason": "length",
  "usage": {
    "input_tokens": 3070,
    "output_tokens": 1024
  }
}

Your request will cost $0.006 per 1000 token.

nvidia/nemotron-3-nano-omni/vision

Input

Result

Logs