TypeError: Failed to fetch

Bug Report
1

I was trying to use nvidia/llama-3.1-nemotron-70b-instruct model with OpenAI API. Using curl everything is working but not in cursor.

I’m getting following error when trying to add OpenAI custom api key:

(status code 0)
TypeError: Failed to fetch

But when running using curl, everything is working fine.

curl https://integrate.api.nvidia.com/v1/chat/completions -H "Content-Type: application/json" -H "Authorization: Bearer nvapi-Zz" -d '{

  "messages": [

    {

      "role": "system",

      "content": "You are a test assistant."

    },

    {

      "role": "user",

      "content": "Testing. Just say hi and nothing else."

    }

  ],

  "model": "nvidia/llama-3.1-nemotron-70b-instruct"

}'

{"id":"chat-dd091bcec726407bb1be66e469a8ab50","object":"chat.completion","created":1729179133,"model":"nvidia/llama-3.1-nemotron-70b-instruct","choices":[{"index":0,"message":{"role":"assistant","content":"Hi."},"logprobs":null,"finish_reason":"stop","stop_reason":null}],"usage":{"prompt_tokens":31,"total_tokens":33,"completion_tokens":2},"prompt_logprobs":null}%
2

I think this has something to do with the endpoint side’s antibot policy.
The cursor uses chromnium underneath, so maybe it’s being blacklisted.

3

How to check what’s the issue ?