Skip to main content
Version: v2

NVIDIA NIM

Generate images using the NVIDIA NIM API. Here, "NIM" refers to two types: running it on your own hardware, or using the API provided by NVIDIA.

Setup Requirements​

  • Endpoint URL: Specify the URL to call the API (e.g., https://ai.api.nvidia.com/v1/genai/black-forest-labs/flux.2-klein-4b). For the web app version, it will be blocked by CORS and won't work, so configure a server, use a CORS proxy, or use the Windows version.
  • Token (Optional): This API key is required unless you are running NIM locally.

Parameter Customization​

Among the parameters written in JSON, if there is no prompt field, the generated prompt will be automatically inserted. If the field exists, the prompt will be inserted before the string in that field. If the field exists and contains the string {prompt}, that part will be replaced with the prompt.

Troubleshooting​

Only Errors Returned​

The image generation model endpoints provided by NVIDIA are not Free Endpoints, so generation may not work, or they may become unusable after a trial number of uses.