Bolt.new AI and OpenAI GPT integration: Step-by-Step Guide 2025

 

Setting Up Environment Variables for OpenAI API

 

• Create an environment variable for your OpenAI API key. Since Bolt.new AI does not have a terminal, add the following code at the beginning of your main file (e.g., index.ts) to set up the variable. Replace YOUROPENAIAPI_KEY with your actual key:

• 
  process.env.OPENAIAPIKEY = "YOUROPENAIAPI_KEY";
      

• This makes the API key available throughout your project.

 

Creating the OpenAI GPT Integration Module

 

• Create a new file called gpt.ts in your project. This file will contain the TypeScript function that calls the OpenAI GPT API.
• Insert the following code into gpt.ts:

• 
  const openaiApiKey = process.env.OPENAIAPIKEY;
  
  export async function getGPTResponse(prompt: string): Promise {
    const endpoint = "https://api.openai.com/v1/chat/completions";
    const headers = {
      "Content-Type": "application/json",
      "Authorization": Bearer ${openaiApiKey}
    };
    const body = {
      model: "gpt-3.5-turbo",
      messages: [{ role: "user", content: prompt }],
      temperature: 0.7
    };
    try {
      const response = await fetch(endpoint, {
        method: "POST",
        headers,
        body: JSON.stringify(body)
      });
      if (!response.ok) {
        throw new Error(Error ${response.status}: ${response.statusText});
      }
      const data = await response.json();
      return data.choices[0].message.content;
    } catch (error) {
      console.error("Error fetching GPT response:", error);
      return "";
    }
  }
      

  
  


• Note: If your runtime does not support the native fetch API (Node.js versions lower than 18), insert the following line at the top of gpt.ts to import the fetch polyfill:
  

  
  import fetch from "node-fetch";
      

  
  Then, ensure that you add the dependency manually in your project’s dependency file if needed.
  

 

Integrating the GPT Module with Your Main Code

 

• In your main file (e.g., index.ts), import the function from gpt.ts by adding the following line near the top:

• 
  import { getGPTResponse } from "./gpt";
      

• Create an endpoint or a function call where you want to use the GPT response. For example, if you have a button click or a form submission, add a function like below:

• 
  async function handleUserPrompt(userPrompt: string) {
    const gptResponse = await getGPTResponse(userPrompt);
    console.log("GPT response:", gptResponse);
    // Further process the response or display it in your UI
  }
      

• Call handleUserPrompt("Your prompt text here") from the appropriate part of your application to trigger the integration.

 

Deploying and Testing Your Integration on Bolt.new AI

 

• Save all your changes in the Bolt.new AI editor.
• No terminal commands are required since Bolt.new AI handles dependencies and deployment automatically.
• Click on the Run button or deploy your project based on your workspace's instructions.
• Trigger the handleUserPrompt function within your application to test if the GPT API responds correctly. Check the console log output for the response from GPT.

 

Additional Notes

 

• Ensure that your OpenAI API key is kept secure and not exposed publicly.
• If you need to add further dependencies (such as node-fetch for older Node versions), follow Bolt.new AI’s instructions for adding dependencies via code comments or project settings since the platform does not support a command-line interface.