Real-Time Chat Application with React and Slash GraphQL

Slash GraphQL is production-ready GraphQL as a Service, which creates a managed GraphQL backend with native GraphQL support using Dgraph so that you can focus on your GraphQL app instead of tables or infrastructure.

In this tutorial, we will learn how to make a real-time chat application, something like the Instagram/Youtube Live chat, where anyone can post a message and everyone else in the chat sees it in real-time. The user just has to spin up a frontend and point it to the same backend as everyone. User management and fancy UI will be out of scope for the sake of simplicity. This app can also be tweaked to make a Hacker News clone, with add-on real-time feature.

Tech Stack

  • Backend: Slash GraphQL
  • Frontend: React + Apollo Client(3.x)

Step 1: Sign Up for Slash GraphQL

If you are a first time user, head on to the Slash GraphQL page and register with your email. You will get an invite in your inbox. Sign-up by following the link in the email and reset the password. This is a one time process.


Step 2: Creating the backend with Slash GraphQL

Login to Slash GraphQL by using the same sign-up credentials and follow the steps below to create a GraphQL backend (images may vary).

Getting Started

Create Backend

Creating Backend

Backend Created

There you go, you have a GraphQL endpoint ready.

Step 3: Create the GraphQL schema

The next step is to create the GraphQL schema. Let’s have a type Message which will consist of id, name, message text, and the timestamp. All fields will be mandatory(enforced by the exclamation mark!). Dgraph auto-generates unique values for scalar ID, think of it like auto-increment primary keys in relational databases.

@withSubscription directive enables subscriptions with the associated type. Subscription queries will not be generated for types without this directive. This way, you can selectively allow/disallow subscriptions on types.

type Message @withSubscription {
  id: ID!
  name: String!
  text: String!
  time: DateTime!

Update Schema

With this, our backend is ready for queries, mutations, and subscriptions.

Step 4: The React App

The frontend is mostly React and Apollo Client boilerplate. The remaining tasks are to establish a GraphQL Subscription on startup, provide the functionality to publish a post, and updating the component on receiving subscription events. Choose the app name of your choice and install these dependencies.

npx create-react-app chatapp

npm install @apollo/client @apollo/link-ws subscriptions-transport-ws graphql

1. index.js

This will be the starting point of the application which create-react-app generates by default.

import React from 'react';
import ReactDOM from 'react-dom';
import App from './App';

    <App />

2. ApolloSetup.js

Next up, create an ApolloClient. The ApolloClient requires a link which defines the protocol and the endpoint to send the requests. For our chat application, we need to support HTTP for queries and mutations and Webscocket for subscriptions. Split helps us to achieve this by accepting both an HTTP and a WebSocket link and uses the appropriate link based on the request. Pass this Split to the ApolloClient. Now, queries and mutations will go over HTTP as normal, but subscriptions will be done over the WebSocket transport.

Set the endpoint variable to the GraphQL endpoint created above. Skip the scheme(https) as it is appended as needed(https/wss).

import { ApolloClient } from '@apollo/client/core';
import { getMainDefinition } from '@apollo/client/utilities';
import { InMemoryCache } from '@apollo/client/cache';
import { HttpLink, split } from '@apollo/client';
import { WebSocketLink } from "@apollo/link-ws";

const endpoint = "<Slash GraphQL endpoint>";

const wsLink = new WebSocketLink({
    uri: `wss://${endpoint}`,
    options: {
        reconnect: true

const httpLink = new HttpLink({
    uri: `https://${endpoint}`

const link = split(
    // split based on operation type
    ({ query }) => {
        const definition = getMainDefinition(query);
        return (
            definition.kind === "OperationDefinition" &&
            definition.operation === "subscription"

export default new ApolloClient({
    cache: new InMemoryCache(),

3. App.js

Enclose the App component with ApolloProvider. This will allow all the enclosed components to use the React Apollo Client features. ApolloProvider requires an ApolloClient to provide the transport features. Import and pass the client created in ApolloSetup.js.

import React from 'react';
import { ApolloProvider } from '@apollo/client'
import Post from './Post';
import apolloClient from './ApolloSetup';

function App() {
  return (
    <ApolloProvider client={apolloClient}>
      <Post />

export default App;

4. Query.js

For this demo, we only have a subscription query for real-time updates and a mutations query for creating a new post.

import { gql } from "@apollo/client";

  subscription {
    queryMessage(order: { desc: time }) {

const SEND_MESSAGE = gql`
  mutation sendMessage($name: String!, $text: String!, $time: DateTime!) {
    addMessage(input: [{ name: $name, text: $text, time: $time }]) {
      message {


5. Post.js

And finally, we have the application logic. For this demo, there is a simple HTML layout that has a form to create new posts, show the total post count, and a table of all the posts made so far by all the users. Set up the apollo client react hooks, one for the mutation and, one for the subscription.

useMutation accepts a mutation query and returns a function on component render, which can be called anytime to execute the mutation.

useSubscription accepts a subscription query and returns a data variable. Every time a subscription event is triggered, the data variable is set with the new data from the backend and the component is re-rendered, thus updating the Post table.

import React, { useState } from "react";
import { useMutation, useSubscription } from "@apollo/client";
import { SUBSCRIPTION_QUERY, SEND_MESSAGE } from "./Query"
import "./index.css";

const Posts = () => {
  const [name, setName] = useState('');
  const [text, setText] = useState('');

  const [sendMessage, { error: mutationError }] = useMutation(SEND_MESSAGE);
  const { data, error: subscriptionError } = useSubscription(SUBSCRIPTION_QUERY);

  if (!data || !data.queryMessage) return (<h1>Connecting...</h1>);
  if (subscriptionError || mutationError) return (<h1>Error...</h1>);

  const handleClick = () => {
    if (name && text) {
      sendMessage({ variables: { name, text, time: (new Date()).toISOString() } });

  return (
      <label>Enter you name : </label>
      <input required type="text" name="name" maxLength="25" onChange={e => setName(}></input>
      <label> Enter your message : </label>
      <input required type="text" name="message" onChange={e => setText(}></input>
      <button type="button" onClick={() => handleClick()}>Post</button>
      <h3>Total Posts : {data.queryMessage.length}</h3>
          { => (
              <td>{(new Date(m.time)).toUTCString()}</td>
              <td align="left">{}</td>
              <td width="1000" align="left">{m.text}</td>

export default Posts;

And with this, the chat app is ready. Run the app as follows. By default, it will run on port 3000.

npm install
npm start

Open http://localhost:3000 in multiple browser tabs and give it a shot. Here is a working demo of the app. The complete source code is hosted here.


To add more features to this app, iterate over the frontend, without worrying about the backend. This is what Slash GraphQL intends to solve. Thank you for following along and hope this was helpful.

Sign-up for Slash GraphQL now and do reach out to us for any queries or feedback.


Chat App code.

Slash GraphQL docs.

React ApolloClient docs.

Cover Photo by NASA on Unsplash

Get blog post to your inbox

Fully Managed GraphQL Backend Service

Introducing Slash GraphQL