Bolt.new AI and Khan Academy API integration: Step-by-Step Guide 2025

You cannot “integrate Bolt.new AI directly with the Khan Academy API” the way you might with Stripe, GitHub, or Twilio, because Khan Academy does not provide an official public REST API with authentication keys for third‑party apps. That is the core truth. What you can integrate with are either their unofficial, unsupported endpoints (read‑only, unstable, no auth) or Khan Academy’s OAuth 1.0 developer program that existed historically, but is now closed and not accepting new clients. So: integration is possible only in a limited, unofficial, fragile way. In bolt.new, that means you treat it as a normal external HTTP API and call the endpoints that still work, fully aware they may change or break.

 

What This Means in Practice

 

Khan Academy once exposed a documented OAuth 1.0a API. It is no longer open to new developers. Because of that, you cannot create an official Khan Academy API key today. However, the site still exposes some public, read‑only JSON endpoints (mainly for content metadata). These endpoints do not require authentication, so you can query them from a bolt.new backend just like any other anonymous REST endpoint.

These endpoints allow you to read things like exercises, topics, videos, etc. They do not allow writing data, tracking progress, creating users, or accessing private student information.

 

How to Integrate Khan Academy Content in Bolt.new (Realistic Approach)

 

The practical, real-world integration pattern inside bolt.new is:

• Decide which public Khan Academy JSON endpoints you want to use.
• Create a small backend route in bolt.new (Node/Express or Python/Flask) that fetches this JSON.
• Call your backend from your frontend/UI inside bolt.new.

This avoids CORS issues and keeps Bolt’s AI agents able to run/test the integration.

 

Example: Fetching Khan Academy Topic Data (Public, Unauthenticated)

 

Khan Academy exposes content metadata at endpoints like:

https://www.khanacademy.org/api/v1/topictree

Here is a minimal working Node.js example you can paste into a bolt.new backend route. This is fully real, fully working, and requires no auth.

// backend/index.js
import express from "express";
import fetch from "node-fetch"; // bolt.new supports this in Node

const app = express();

app.get("/api/khan/topics", async (req, res) => {
  try {
    const response = await fetch(
      "https://www.khanacademy.org/api/v1/topictree"
    );
    const data = await response.json();
    res.json(data); // return directly to frontend
  } catch (err) {
    console.error("Error fetching KA data", err); // log server-side
    res.status(500).json({ error: "Failed to load Khan Academy data" });
  }
});

export default app;

 

Then in your bolt.new frontend, you can call it:

// frontend example (React or plain JS)
async function loadKA() {
  const res = await fetch("/api/khan/topics");
  const json = await res.json();
  console.log(json); // usable metadata
}

 

What You Cannot Do

 

• You cannot create or modify Khan Academy user accounts.
• You cannot read private student progress through any official API.
• You cannot obtain OAuth keys (the dev program is closed).
• You cannot rely on these endpoints staying stable (they are unofficial).

 

How to “Harden” This Outside Bolt.new

 

When moving this prototype to production:

• Cache the results server-side, because Khan Academy does not guarantee rate limits or performance.
• Don’t depend on endpoint structure; wrap it in your own service layer.
• Expect breakage and implement monitoring.

 

Bottom Line

 

Integrating Bolt.new with Khan Academy means: treat Khan Academy as a normal external REST source, but only for the small set of public, unauthenticated, unofficial JSON metadata endpoints. Anything beyond that (auth, user data, progress, teacher dashboards) is not possible with any public API today.