This guide walks you through:

Creating a project
Securing your secret key
Creating an execution endpoint
Integrating FusePlane with your app
Calling endpoints from the frontend

⏱️ Average setup time: 5–10 minutes

Create a Project

Step 1: Create a new project

From the FusePlane dashboard:

Click New Project
Give it a name
Choose environment (development or production)

Step 2: Get credentials

Once created, you'll receive:

Secret Project Key
Base Endpoint URL

Project Credentials

Secret Key:eb_sk_live_xxxxxxxxx

Base URL:https://fuseplane.com/api/p

Important

Your Secret Project Key must never be exposed in the frontend. It is only used inside a server environment.

Set Up the Server Bundler (Key Security Layer)

To securely attach your FusePlane secret key, you need a lightweight server route. This route simply forwards requests to FusePlane.

Choose one of the integrations below:

ANext.js (App Router)

Recommended

1. Create Route Handler

/app/api/p/[...path]/route.ts

export async function ALL(req, { params }) {
  const path = params.path.join("/")
  const url = `${process.env.FUSEPLANE_BASE_URL}/${path}`

  const response = await fetch(url, {
    method: req.method,
    headers: {
      "Authorization": `Bearer ${process.env.FUSEPLANE_SECRET_KEY}`,
      "Content-Type": req.headers.get("content-type") || "application/json"
    },
    body: req.method === "GET" ? undefined : await req.text()
  })

  return new Response(response.body, {
    status: response.status,
    headers: response.headers
  })
}

2. Environment Variables (.env)

FUSEPLANE_SECRET_KEY=eb_sk_live_xxxxx
FUSEPLANE_BASE_URL=https://fuseplane.com/api/p

BVercel Serverless Function

File structure:

/api/p/[...path].js

Handler:

export default async function handler(req, res) {
  const path = req.query.path.join("/")
  const url = `${process.env.FUSEPLANE_BASE_URL}/${path}`

  const response = await fetch(url, {
    method: req.method,
    headers: {
      Authorization: `Bearer ${process.env.FUSEPLANE_SECRET_KEY}`,
      "Content-Type": "application/json"
    },
    body: req.method === "GET" ? null : JSON.stringify(req.body)
  })

  const data = await response.text()
  res.status(response.status).send(data)
}

CNetlify Functions

File structure:

/netlify/functions/p.js

Handler:

export async function handler(event) {
  const path = event.path.split("/p/")[1]
  const url = `${process.env.FUSEPLANE_BASE_URL}/${path}`

  const response = await fetch(url, {
    method: event.httpMethod,
    headers: {
      Authorization: `Bearer ${process.env.FUSEPLANE_SECRET_KEY}`,
      "Content-Type": "application/json"
    },
    body: event.httpMethod === "GET" ? null : event.body
  })

  return {
    statusCode: response.status,
    body: await response.text()
  }
}

Create an Execution Endpoint (API Gateway)

Now you'll define what your frontend is allowed to call.

Go to: Project Dashboard → API Gateway → Create Endpoint

Step 1: Name & Capability

• Endpoint Name: (e.g. chat-completion)
• Capability: Choose based on API type (AI, Payments, Messaging, Generic HTTP, etc.)

This helps FusePlane configure correct security behavior.

Step 2: Upstream Destination

This is the original API URL you were planning to call directly from the frontend.

https://api.openai.com/v1/chat/completions

💡

Tip: You can paste a full fetch() request, OR describe what you're building. FusePlane will suggest the correct upstream URL.

Warning: If the upstream URL is incorrect, the response will also be incorrect. Always verify this step carefully.

Step 3: Key Configuration (Auto-Validation)

FusePlane analyzes the upstream API and shows required headers / keys.

Authorization: Bearer {{OPENAI_API_KEY}}

You must:

Create the keys exactly as suggested
Map them to provider secrets you added earlier

FusePlane validates this configuration automatically.

Step 4: Create & Deploy

Click Create Endpoint

Your endpoint is:

validated
secured
deployed instantly

Example deployed path:

POST /openai/chat

Calling the Endpoint from the Frontend

Now your frontend talks only to your app's API route.

Frontend Example (React / Next.js)

const response = await fetch("/api/p/openai/chat", {
  method: "POST",
  headers: {
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
    model: "gpt-4",
    messages: [
      { role: "user", content: "Hello!" }
    ]
  })
})

const data = await response.json()
console.log(data)

You're all set!

Your frontend now securely calls third-party APIs through FusePlane without exposing any secrets.

Ready to build?

Your integration is complete. Head to the dashboard to manage your endpoints and monitor usage.

API Reference