NVIDIA_API_KEY access

I’m trying to follow this guide and I keep getting the same error.
It says To get started, create a free account with the NVIDIA API catalog and follow these steps:

  1. Select any model.
  2. Choose Python, Get API Key .
  3. Save the generated key as NVIDIA_API_KEY.

From there, you should have access to the endpoints.

This is exactly what I do and I keep getting this error:

Exception: [401] Unauthorized
invalid response from UAM
Please check or regenerate your API key.

When I run the provided sample code:
from langchain_nvidia_ai_endpoints import ChatNVIDIA

client = ChatNVIDIA(
  model="meta/llama-3.1-405b-instruct",
  api_key=api_key, 
  temperature=0.2,
  top_p=0.7,
  max_tokens=1024,
)

for chunk in client.stream([{"role":"user","content":"Write a limerick about the wonders of GPU computing."}]): 
  print(chunk.content, end="")

Hi! Have you found a fix for this issue? Got the same problem and I can’t think of a solution

Please try generating a new API Key and using that.

what’s the the term of validity of the api key? I found that it often failed in another day.

1 Like

Generating new API key doen’t solve it

בתאריך יום ה׳, 9 בינו׳ 2025, 13:22, מאת Pengyiyu2012 via NVIDIA Developer Forums ‏<[email protected]>:

any update on this new api key is throwing 403 after ~24 hours, is there any validity of a api key?

Sorry to hear this - have passed on your feedback to the team who are working to improve the experience.

Hello! Any progress on this issue? We have been experiencing the same thing.

Is it not becasue of the varible misnaming (at least appears to be)? it looks as if you name the api key variable “NVIDIA_API_KEY” whern you make the key, then it looks like you are calling it in like “api_key=api_key,”
?
try maybe api_key=NVIDIA_API_KEY ?