#Can't use Nvidia's model from openrouter

3 messages · Page 1 of 1 (latest)

oak glen
#

WIth Openclaw, I want to use nvidia's model: nvidia/nemotron-3-super-120b-a12b:free from openrouter, but it didn't suggest in model while onboarding. also tried with custom name model set with this, but not working

timid laurelBOT
#

Let me verify the docs and the actual model wiring so I don’t guess the wrong key/path.

#

Yep, I think the issue is the model ref.

nvidia/nemotron-3-super-120b-a12b:free gets parsed as the native NVIDIA provider in OpenClaw, not OpenRouter.

If you're using an OpenRouter key, use this instead:

openclaw models set openrouter/nvidia/nemotron-3-super-120b-a12b:free
openclaw models status

A couple notes:

  • Onboarding for OpenRouter defaults to openrouter/auto, it does not suggest every OpenRouter model there.
  • OpenRouter model refs need the full form: openrouter/<provider>/<model>

If you want to confirm the exact current OpenRouter ID first: