Hi NVIDIA team,
My newly generated nvapi- key authenticates correctly but is rejected on all inference calls with a 403 “Authorization failed”. Based on other threads in this category, I believe my Personal Organization is missing the “Public API Endpoints” permission, which I can’t enable myself.
Account details:
- Account email:
suleymanabdirahman@gmail.com - Organization:
suley
What works:
GET https://integrate.api.nvidia.com/v1/models→ HTTP 200 (returns the full model catalog, so the key authenticates fine)
What fails:
POST https://integrate.api.nvidia.com/v1/chat/completions→ HTTP 403- Response body:
{"status": 403, "title": "Forbidden", "detail": "Authorization failed"} - This happens on every model I try, including
meta/llama-3.1-8b-instruct,nvidia/llama-3.1-nemotron-nano-8b-v1, andnvidia/llama-3.3-nemotron-super-49b-v1.5— so it is not model-specific.
Since it’s a 403 (not a 401) and listing models succeeds, this looks like an authorization/permission gap rather than a bad key.
Please enable the “Public API Endpoints” permission for my Personal Organization (suley) so I can make inference calls with my existing key.
Thank you!