Error with nvidia api key

[backupafelix045-boop]

Not Found: 404 page not found

Open Design run error diagnostics
trace_id: 2a13cf21-dc67-4362-a5fd-3b16259ea138
run_id: 2a13cf21-dc67-4362-a5fd-3b16259ea138
error_code: AGENT_EXECUTION_FAILED
project_id: 75bad2be-a62c-4bbb-8b76-aa04a839b566
conversation_id: b31c3119-d155-4657-9d60-d007bda91462
assistant_message_id: ce69a54c-159f-4a51-873a-127bedc4c9ed
agent_id: openai-api

I get this error whenever I try to use the chat or create a design. I'm not sure if I have something misconfigured, but I need help. I can't create absolutely anything. Does anyone know what the problem is?

And could you explain it to me as if I were a five-year-old?

I'm using a free NVIDIA API key, but it doesn't seem to be working for me.

[roian6]

I traced the exact error text rather than guessing from the API key.

Five-year-old version: NVIDIA's API has a front door and a specific room. Your request appears to be reaching the building, but not the /v1 room.

A request to:

https://integrate.api.nvidia.com/chat/completions

currently returns the same plain 404 page not found text shown in your screenshot. NVIDIA's OpenAI-compatible Base URL should be:

https://integrate.api.nvidia.com/v1

Could you check these three things?

1. The provider/protocol is OpenAI-compatible.
2. The Base URL is exactly https://integrate.api.nvidia.com/v1.
3. The model was selected from the model list returned for your NVIDIA account, rather than copied from an older example.

Then retry the connection and one small chat request.

If it still fails, please share the Base URL, model ID, and the redacted error body — but do not post your API key.

There is also a possible Open Design validation gap here: the connection-test path can add /v1 while probing a custom OpenAI-compatible endpoint, but the actual agent runtime preserves a custom NVIDIA URL as entered. If adding /v1 fixes your run, that would confirm the mismatch rather than a bad key.