Simple request + error which occurs before streaming begins (where applicable).

-- request --
{
  "messages": [
    {
      "role": "user",
      "content": "how many angels can dance on the head of a pin\n"
    }
  ],
  "model": "gpt-4.1",
  "stream": true
}

-- streaming --
HTTP/2.0 400 Bad Request
Content-Length: 281
Content-Type: application/json

{
  "error": {
    "message": "Input tokens exceed the configured limit of 272000 tokens. Your messages resulted in 3148588 tokens. Please reduce the length of the messages.",
    "type": "invalid_request_error",
    "param": "messages",
    "code": "context_length_exceeded"
  }
}


-- non-streaming --
HTTP/2.0 400 Bad Request
Content-Length: 281
Content-Type: application/json

{
  "error": {
    "message": "Input tokens exceed the configured limit of 272000 tokens. Your messages resulted in 3148588 tokens. Please reduce the length of the messages.",
    "type": "invalid_request_error",
    "param": "messages",
    "code": "context_length_exceeded"
  }
}