ShareGPT

Share your wildest ChatGPT conversations with one click.

Introduction · Tech Stack · REST API

Introduction

ShareGPT is an open-source Chrome Extension for you to share your wildest ChatGPT conversations with one click.

Features

• Share your ChatGPT conversations with one-click
• Browse examples on sharegpt.com/explore
• Save your favorite conversations for later
• Leave comments on conversations

Tech Stack

ShareGPT is built with the following stack:

• Next.js – framework
• Typescript – language
• Tailwind – CSS
• Upstash – redis
• Planetscale – database
• NextAuth.js – auth
• Vercel – hosting

REST API

The ShareGPT API is a REST-styled API that allows you to write and read conversations from our database, exposed as HTTP endpoints.

Conversations Endpoint

POST: https://sharegpt.com/api/conversations

You can use this endpoint to add new conversations to our database.

First, if you haven't already, process the ShareGPT conversation using the following code:

function conversationData() {
  const threadContainer = document.querySelector(
    "#__next main div:nth-of-type(1) div:nth-of-type(1) div:nth-of-type(1) div:nth-of-type(1)"
  );

  var result = {
    avatarUrl: getAvatarImage(),
    items: [],
  };

  for (const node of threadContainer.children) {
    const markdownContent = node.querySelector(".markdown");

    // tailwind class indicates human or gpt
    if ([...node.classList].includes("dark:bg-gray-800")) {
      result.items.push({
        from: "human",
        value: node.textContent,
      });
      // if it's a GPT response, it might contain code blocks
    } else if ([...node.classList].includes("bg-gray-50")) {
      result.items.push({
        from: "gpt",
        value: markdownContent.outerHTML,
      });
    }
  }

  return result;
}

function getAvatarImage() {
  // Create a canvas element
  const canvas = document.createElement("canvas");

  const image = document.querySelectorAll("img")[1];

  // Set the canvas size to 30x30 pixels
  canvas.width = 30;
  canvas.height = 30;

  // Draw the img onto the canvas
  canvas.getContext("2d").drawImage(image, 0, 0);

  // Convert the canvas to a base64 string as a JPEG image
  const base64 = canvas.toDataURL("image/jpeg");

  return base64;
}

Then, send a POST request to the endpoint above with the following payload and request headers:

const res = await fetch("https://sharegpt.com/api/conversations", {
  body: JSON.stringify(conversationData),
  headers: {
    "Content-Type": "application/json",
  },
  method: "POST",
});

This will return an object with an id attribute which will be the unique identifier for the generated post:

const { id } = await res.json();
const url = `https://shareg.pt/${id}`; // short link to the ShareGPT post

GET: https://sharegpt.com/api/conversations

This endpoint takes 3 optional query parameters:

• type:
• Used for sorting the results.
• Takes 2 string values: "new" | "top"
• "new" sorts conversations by creation time
• "top" sorts conversations by number of views
• If undefined, defaults to "top"
• page:
• Used for pagination
• Takes an integer value as a factor of the PAGINATION_LIMIT, which is set to 50.
• E.g. to get posts 100 - 150, set page to 3
• If undefined, defaults to 1
• search
• Used for filtering records by title.
• E.g. search = "python" returns all records with the word "python" in the title
• If undefined, search results are not filtered

Example:

await fetch(
  "https://sharegpt.com/api/conversations?type=new&page=2&search=python"
);

This returns a list of conversations with the following type:

interface ConversationMeta {
  id: string; // unique id for the conversation
  title: string; // title of the conversation (first user prompt)
  avatar: string; // base64 encoded URI of the user's avatar
  saves: number; // number of times the conversation is saved on ShareGPT
  comments: number; // number of comments the conversation has on ShareGPT
  views: number; // number of times the conversation has been viewed on ShareGPT
  createdAt: Date; // timestamp when the conversation was creataed
}
[];