Incorrectly getting error 429 (1015) with Dall-E 3

API
,
1

I am running a deployment on vercel which is using the Dalle-3 and completions APIs. The completion API works correctly, but ever since last morning, I have been getting error 429 (1015) with dalle-3 generation from my vercel deployment. Using the same key however, I can generate images on my development server.

I definitely have enough money in my account and I am only making a couple requests per day. Trying a new API key did not work either.

This is very similar to Error 429 (1015) with Dall-E 3, but quota/rate not reached (Bubble related?) but I am not using any third party in this case.

2

Here’s fun:

Insert into your code, right before the API call, a printout or log of the environment variable OPENAI_API_KEY, and then also a dump of api_key from the client of the library module.

See if the same results are reported between working environment and failing deployment.

3

I’m not sure exactly what you mean, but yes, both on my development server and deployed vercel app, both are logged as the same key.

"use server";

import OpenAI from "openai";

const openai = new OpenAI({
  apiKey: process.env.OPENAI_KEY,
});

async function generateImage(prompt: string): Promise<string> {
  console.log(process.env.OPENAI_KEY); // the same
  console.log(openai.apiKey); // the same
  const imageResponse = await openai.images.generate({
    prompt,
    model: "dall-e-3",
    quality: "standard",
    response_format: "b64_json",
    size: "1024x1024",
    style: "vivid",
    n: 1,
  });

  return imageResponse.data[0].b64_json as string;
}