OpenAI API error: "TypeError: Cannot read properties of undefined (reading 'choices')"

Question

I looked into various possible solutions:

• OpenAI and Javascript error : Getting 'TypeError: Cannot read properties of undefined (reading 'create') at Object.<anonymous>"
• https://github.com/davila7/code-gpt-docs/issues/57
• https://www.npmjs.com/package/openai

Here is my code:

import { config } from "dotenv"
config()

// New
import OpenAI from 'openai';
import readline from "readline"

const openai = new OpenAI({
  apiKey: 'My Key', // defaults to process.env["OPENAI_API_KEY"]
});

const userInterface = new readline.createInterface({
    input: process.stdin,
    output: process.stdout

})

userInterface.prompt()
userInterface.on("line", async input => {
    const res = await openai.chat.completions.create({
        model: 'gpt-3.5-turbo',
        messages: [{ role: 'user', content: input}],
    })
    console.log(res.data.choices[0].message.content)
    userInterface.prompt()
})

When I run the code, I get the error message from the title of the post: TypeError: Cannot read properties of undefined (reading 'choices').

I tried the following to fix the issue:

• Reload the Window with command: Developer:Reload.Window
• Using different OpenAI API keys
• Update package.json OpenAI to the latest version which is 4.11.1
• Generally a different console.log code/output & code structure to solve the issue

Can anyone point me in the right direction? Everything works, besides the user input. As a guideline I used this video: https://www.youtube.com/watch?v=4qNwoAAfnk4. I updated the project as far as I could. Why does the user input not work?

Also, the code on my PC has a functioning apiKey implemented. Thanks for reading!

Answer 1

Changing console.log(res.data.choices[0].message.content) to console.log(res.choices[0]) or console.log(res.choices[0].message.content) solved the issue and fixed the proper user input feature.

Answer 2

OpenAI NodeJS SDK v4 was released on August 16, 2023, and is a complete rewrite of the SDK. Among other things, there are changes in extracting the message content. See the v3 to v4 migration guide.

• If you have the OpenAI NodeJS SDK v3:

import { Configuration, OpenAIApi } from 'openai';

const configuration = new Configuration({
  apiKey: process.env.OPENAI_API_KEY,
});

const openai = new OpenAIApi(configuration);

const chatCompletion = await openai.createChatCompletion({
  model: 'gpt-3.5-turbo',
  messages: [{ role: 'user', content: 'Hello!' }],
});

console.log(chatCompletion.data.choices[0].message);

• If you have the OpenAI NodeJS SDK v4:

import OpenAI from 'openai';

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

const chatCompletion = await openai.chat.completions.create({
  model: 'gpt-3.5-turbo',
  messages: [{ role: 'user', content: 'Hello!'}],
});

console.log(chatCompletion.choices[0].message); // There's no "data" property