Building a ChatGPT Clone with Supabase Edge Functions and OpenAI

13. June 2025 approx. 46 min read Supabase

Contents 61

Build a fully functional ChatGPT clone using Supabase Edge Functions, OpenAI's API, and React. This comprehensive guide covers everything from authentication to real-time streaming responses.

Introduction
Prerequisites
Project Setup
Database Schema
Edge Functions Setup
Building the Chat Interface
Implementing Streaming Responses
Adding Conversation History
Performance Optimization
Security Best Practices
Deployment
Conclusion

Introduction

In this tutorial, we'll build a production-ready ChatGPT clone that features:

Real-time streaming responses like ChatGPT
Persistent conversation history
User authentication
Rate limiting and usage tracking
Mobile-responsive design
Markdown rendering with syntax highlighting

By leveraging Supabase Edge Functions, we can create a serverless API that scales automatically and provides low-latency responses globally. Combined with OpenAI's powerful language models, we'll create an experience that rivals commercial chat applications.

Prerequisites

Before we start, make sure you have:

A Supabase account (free tier works fine)
An OpenAI API key
Node.js 18+ installed
Basic knowledge of React and TypeScript
Familiarity with SQL

Project Setup

Step 1: Create a New Supabase Project

Head over to database.new to create a new Supabase project. Save your project URL and anon key - we'll need these later.

Step 2: Initialize Your React App

We'll use Vite for a fast development experience:

npm create vite@latest chatgpt-clone -- --template react-ts
cd chatgpt-clone
npm install

Install the required dependencies:

npm install @supabase/supabase-js @supabase/auth-ui-react @supabase/auth-ui-shared
npm install react-markdown remark-gfm react-syntax-highlighter
npm install @types/react-syntax-highlighter --save-dev

Step 3: Set Up Environment Variables

Create a .env file in your project root:

VITE_SUPABASE_URL=your_supabase_project_url
VITE_SUPABASE_ANON_KEY=your_supabase_anon_key

Step 4: Initialize Supabase Client

Create src/lib/supabase.ts:

import { createClient } from '@supabase/supabase-js'

const supabaseUrl = import.meta.env.VITE_SUPABASE_URL
const supabaseAnonKey = import.meta.env.VITE_SUPABASE_ANON_KEY

export const supabase = createClient(supabaseUrl, supabaseAnonKey)

Database Schema

Let's set up our database tables to store conversations and messages. Navigate to the SQL Editor in your Supabase dashboard and run:

-- Enable UUID extension
CREATE EXTENSION IF NOT EXISTS "uuid-ossp";

-- Create conversations table
CREATE TABLE conversations (
  id UUID DEFAULT uuid_generate_v4() PRIMARY KEY,
  user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE,
  title TEXT NOT NULL,
  created_at TIMESTAMPTZ DEFAULT NOW(),
  updated_at TIMESTAMPTZ DEFAULT NOW()
);

-- Create messages table
CREATE TABLE messages (
  id UUID DEFAULT uuid_generate_v4() PRIMARY KEY,
  conversation_id UUID REFERENCES conversations(id) ON DELETE CASCADE,
  role TEXT NOT NULL CHECK (role IN ('user', 'assistant', 'system')),
  content TEXT NOT NULL,
  created_at TIMESTAMPTZ DEFAULT NOW()
);

-- Create usage_stats table for tracking API usage
CREATE TABLE usage_stats (
  id UUID DEFAULT uuid_generate_v4() PRIMARY KEY,
  user_id UUID REFERENCES auth.users(id) ON DELETE CASCADE,
  tokens_used INTEGER NOT NULL,
  model TEXT NOT NULL,
  created_at TIMESTAMPTZ DEFAULT NOW()
);

-- Add indexes for better performance
CREATE INDEX idx_conversations_user_id ON conversations(user_id);
CREATE INDEX idx_messages_conversation_id ON messages(conversation_id);
CREATE INDEX idx_usage_stats_user_id ON usage_stats(user_id);

-- Enable Row Level Security (RLS)
ALTER TABLE conversations ENABLE ROW LEVEL SECURITY;
ALTER TABLE messages ENABLE ROW LEVEL SECURITY;
ALTER TABLE usage_stats ENABLE ROW LEVEL SECURITY;

-- RLS Policies for conversations
CREATE POLICY "Users can view own conversations" ON conversations
  FOR SELECT USING (auth.uid() = user_id);

CREATE POLICY "Users can create own conversations" ON conversations
  FOR INSERT WITH CHECK (auth.uid() = user_id);

CREATE POLICY "Users can update own conversations" ON conversations
  FOR UPDATE USING (auth.uid() = user_id);

CREATE POLICY "Users can delete own conversations" ON conversations
  FOR DELETE USING (auth.uid() = user_id);

-- RLS Policies for messages
CREATE POLICY "Users can view messages from own conversations" ON messages
  FOR SELECT USING (
    EXISTS (
      SELECT 1 FROM conversations
      WHERE conversations.id = messages.conversation_id
      AND conversations.user_id = auth.uid()
    )
  );

CREATE POLICY "Users can create messages in own conversations" ON messages
  FOR INSERT WITH CHECK (
    EXISTS (
      SELECT 1 FROM conversations
      WHERE conversations.id = messages.conversation_id
      AND conversations.user_id = auth.uid()
    )
  );

-- RLS Policies for usage_stats
CREATE POLICY "Users can view own usage stats" ON usage_stats
  FOR SELECT USING (auth.uid() = user_id);

CREATE POLICY "Service role can insert usage stats" ON usage_stats
  FOR INSERT WITH CHECK (true);

-- Function to update conversation timestamp
CREATE OR REPLACE FUNCTION update_conversation_timestamp()
RETURNS TRIGGER AS $$
BEGIN
  UPDATE conversations
  SET updated_at = NOW()
  WHERE id = NEW.conversation_id;
  RETURN NEW;
END;
$$ LANGUAGE plpgsql;

-- Trigger to update conversation timestamp on new message
CREATE TRIGGER update_conversation_on_message
AFTER INSERT ON messages
FOR EACH ROW
EXECUTE FUNCTION update_conversation_timestamp();

Edge Functions Setup

Now let's create our Edge Function to handle chat completions. This function will communicate with OpenAI's API and stream responses back to the client.

Step 1: Install Supabase CLI

npm install -g supabase

Step 2: Initialize Supabase Functions

In your project root:

supabase init
supabase functions new chat-completion

Step 3: Create the Chat Completion Function

Edit supabase/functions/chat-completion/index.ts:

import { serve } from "https://deno.land/std@0.168.0/http/server.ts"
import { createClient } from 'https://esm.sh/@supabase/supabase-js@2'

const corsHeaders = {
  'Access-Control-Allow-Origin': '*',
  'Access-Control-Allow-Headers': 'authorization, x-client-info, apikey, content-type',
}

interface ChatRequest {
  messages: Array<{
    role: 'user' | 'assistant' | 'system'
    content: string
  }>
  conversationId?: string
  temperature?: number
  model?: string
}

serve(async (req) => {
  // Handle CORS
  if (req.method === 'OPTIONS') {
    return new Response('ok', { headers: corsHeaders })
  }

  try {
    const supabaseClient = createClient(
      Deno.env.get('SUPABASE_URL') ?? '',
      Deno.env.get('SUPABASE_SERVICE_ROLE_KEY') ?? '',
      { 
        global: { 
          headers: { Authorization: req.headers.get('Authorization')! }
        }
      }
    )

    // Get the current user
    const {
      data: { user },
    } = await supabaseClient.auth.getUser()

    if (!user) {
      return new Response(
        JSON.stringify({ error: 'Unauthorized' }),
        { 
          status: 401,
          headers: { ...corsHeaders, 'Content-Type': 'application/json' }
        }
      )
    }

    const { messages, conversationId, temperature = 0.7, model = 'gpt-3.5-turbo' } = await req.json() as ChatRequest

    // Create or verify conversation
    let convId = conversationId
    if (!convId) {
      const { data: newConversation, error: convError } = await supabaseClient
        .from('conversations')
        .insert({ 
          user_id: user.id, 
          title: messages[0]?.content.substring(0, 50) + '...' || 'New Chat'
        })
        .select()
        .single()

      if (convError) throw convError
      convId = newConversation.id
    }

    // Save user message
    const { error: msgError } = await supabaseClient
      .from('messages')
      .insert({
        conversation_id: convId,
        role: 'user',
        content: messages[messages.length - 1].content
      })

    if (msgError) throw msgError

    // Create OpenAI API request
    const openAiResponse = await fetch('https://api.openai.com/v1/chat/completions', {
      method: 'POST',
      headers: {
        'Authorization': `Bearer ${Deno.env.get('OPENAI_API_KEY')}`,
        'Content-Type': 'application/json',
      },
      body: JSON.stringify({
        model,
        messages,
        temperature,
        stream: true,
      }),
    })

    if (!openAiResponse.ok) {
      const error = await openAiResponse.json()
      throw new Error(`OpenAI API error: ${error.error?.message || 'Unknown error'}`)
    }

    // Set up SSE stream
    const encoder = new TextEncoder()
    const decoder = new TextDecoder()

    let fullResponse = ''

    const stream = new ReadableStream({
      async start(controller) {
        const reader = openAiResponse.body?.getReader()
        if (!reader) return

        try {
          while (true) {
            const { done, value } = await reader.read()
            if (done) break

            const chunk = decoder.decode(value)
            const lines = chunk.split('\n')
            
            for (const line of lines) {
              if (line.startsWith('data: ')) {
                const data = line.slice(6)
                if (data === '[DONE]') {
                  // Save assistant response to database
                  await supabaseClient
                    .from('messages')
                    .insert({
                      conversation_id: convId,
                      role: 'assistant',
                      content: fullResponse
                    })

                  // Track usage
                  const usage = {
                    user_id: user.id,
                    tokens_used: Math.ceil(fullResponse.length / 4), // Rough estimate
                    model
                  }
                  
                  await supabaseClient
                    .from('usage_stats')
                    .insert(usage)

                  controller.enqueue(encoder.encode(`data: {"done": true, "conversationId": "${convId}"}\n\n`))
                  controller.close()
                  return
                }

                try {
                  const parsed = JSON.parse(data)
                  const content = parsed.choices[0]?.delta?.content || ''
                  fullResponse += content
                  
                  if (content) {
                    controller.enqueue(encoder.encode(`data: ${JSON.stringify({ content })}\n\n`))
                  }
                } catch (e) {
                  console.error('Error parsing stream:', e)
                }
              }
            }
          }
        } catch (error) {
          controller.error(error)
        }
      },
    })

    return new Response(stream, {
      headers: {
        ...corsHeaders,
        'Content-Type': 'text/event-stream',
        'Cache-Control': 'no-cache',
        'Connection': 'keep-alive',
      },
    })

  } catch (error) {
    console.error('Error:', error)
    return new Response(
      JSON.stringify({ error: error.message }),
      { 
        status: 500,
        headers: { ...corsHeaders, 'Content-Type': 'application/json' }
      }
    )
  }
})

Step 4: Add OpenAI API Key

In your Supabase dashboard, navigate to Edge Functions > Secrets and add:

Key: OPENAI_API_KEY
Value: Your OpenAI API key

Step 5: Deploy the Edge Function

supabase functions deploy chat-completion

Building the Chat Interface

Now let's create the React components for our chat interface.

Step 1: Create the Auth Component

Create src/components/Auth.tsx:

import { Auth } from '@supabase/auth-ui-react'
import { ThemeSupa } from '@supabase/auth-ui-shared'
import { supabase } from '../lib/supabase'

export default function AuthComponent() {
  return (
    <div className="auth-container">
      <h1>ChatGPT Clone</h1>
      <p>Sign in to start chatting</p>
      <Auth
        supabaseClient={supabase}
        appearance={{ theme: ThemeSupa }}
        providers={['google', 'github']}
        redirectTo={window.location.origin}
      />
    </div>
  )
}

Step 2: Create the Message Component

Create src/components/Message.tsx:

import ReactMarkdown from 'react-markdown'
import remarkGfm from 'remark-gfm'
import { Prism as SyntaxHighlighter } from 'react-syntax-highlighter'
import { vscDarkPlus } from 'react-syntax-highlighter/dist/esm/styles/prism'

interface MessageProps {
  role: 'user' | 'assistant' | 'system'
  content: string
}

export default function Message({ role, content }: MessageProps) {
  return (
    <div className={`message ${role}`}>
      <div className="message-role">
        {role === 'user' ? '👤' : '🤖'} {role}
      </div>
      <div className="message-content">
        <ReactMarkdown
          remarkPlugins={[remarkGfm]}
          components={{
            code({ node, inline, className, children, ...props }) {
              const match = /language-(\w+)/.exec(className || '')
              return !inline && match ? (
                <SyntaxHighlighter
                  style={vscDarkPlus}
                  language={match[1]}
                  PreTag="div"
                  {...props}
                >
                  {String(children).replace(/\n$/, '')}
                </SyntaxHighlighter>
              ) : (
                <code className={className} {...props}>
                  {children}
                </code>
              )
            }
          }}
        >
          {content}
        </ReactMarkdown>
      </div>
    </div>
  )
}

Step 3: Create the Chat Component

Create src/components/Chat.tsx:

import { useState, useEffect, useRef } from 'react'
import { supabase } from '../lib/supabase'
import Message from './Message'
import { User } from '@supabase/supabase-js'

interface ChatMessage {
  role: 'user' | 'assistant' | 'system'
  content: string
}

interface Conversation {
  id: string
  title: string
  created_at: string
  updated_at: string
}

interface ChatProps {
  user: User
}

export default function Chat({ user }: ChatProps) {
  const [messages, setMessages] = useState<ChatMessage[]>([])
  const [input, setInput] = useState('')
  const [isLoading, setIsLoading] = useState(false)
  const [conversations, setConversations] = useState<Conversation[]>([])
  const [currentConversation, setCurrentConversation] = useState<string | null>(null)
  const [streamingMessage, setStreamingMessage] = useState('')
  const messagesEndRef = useRef<HTMLDivElement>(null)
  const abortControllerRef = useRef<AbortController | null>(null)

  useEffect(() => {
    loadConversations()
  }, [])

  useEffect(() => {
    scrollToBottom()
  }, [messages, streamingMessage])

  useEffect(() => {
    if (currentConversation) {
      loadMessages(currentConversation)
    }
  }, [currentConversation])

  const scrollToBottom = () => {
    messagesEndRef.current?.scrollIntoView({ behavior: 'smooth' })
  }

  const loadConversations = async () => {
    const { data, error } = await supabase
      .from('conversations')
      .select('*')
      .order('updated_at', { ascending: false })

    if (!error && data) {
      setConversations(data)
      if (data.length > 0 && !currentConversation) {
        setCurrentConversation(data[0].id)
      }
    }
  }

  const loadMessages = async (conversationId: string) => {
    const { data, error } = await supabase
      .from('messages')
      .select('*')
      .eq('conversation_id', conversationId)
      .order('created_at', { ascending: true })

    if (!error && data) {
      setMessages(data.map(msg => ({
        role: msg.role,
        content: msg.content
      })))
    }
  }

  const handleSubmit = async (e: React.FormEvent) => {
    e.preventDefault()
    if (!input.trim() || isLoading) return

    const userMessage = input.trim()
    setInput('')
    setIsLoading(true)
    setStreamingMessage('')

    // Add user message to UI
    const newMessages = [...messages, { role: 'user' as const, content: userMessage }]
    setMessages(newMessages)

    try {
      // Create abort controller for this request
      abortControllerRef.current = new AbortController()

      const { data: { session } } = await supabase.auth.getSession()
      
      const response = await fetch(
        `${import.meta.env.VITE_SUPABASE_URL}/functions/v1/chat-completion`,
        {
          method: 'POST',
          headers: {
            'Content-Type': 'application/json',
            'Authorization': `Bearer ${session?.access_token}`
          },
          body: JSON.stringify({
            messages: newMessages,
            conversationId: currentConversation,
            model: 'gpt-3.5-turbo',
            temperature: 0.7
          }),
          signal: abortControllerRef.current.signal
        }
      )

      if (!response.ok) {
        throw new Error('Failed to get response')
      }

      const reader = response.body?.getReader()
      const decoder = new TextDecoder()
      let assistantMessage = ''

      if (reader) {
        while (true) {
          const { done, value } = await reader.read()
          if (done) break

          const chunk = decoder.decode(value)
          const lines = chunk.split('\n')

          for (const line of lines) {
            if (line.startsWith('data: ')) {
              try {
                const data = JSON.parse(line.slice(6))
                
                if (data.done) {
                  // Update conversation ID if new
                  if (data.conversationId && !currentConversation) {
                    setCurrentConversation(data.conversationId)
                    loadConversations()
                  }
                  
                  // Add complete message to messages
                  setMessages(prev => [...prev, { 
                    role: 'assistant' as const, 
                    content: assistantMessage 
                  }])
                  setStreamingMessage('')
                } else if (data.content) {
                  assistantMessage += data.content
                  setStreamingMessage(assistantMessage)
                }
              } catch (e) {
                console.error('Error parsing stream:', e)
              }
            }
          }
        }
      }
    } catch (error: any) {
      if (error.name === 'AbortError') {
        console.log('Request aborted')
      } else {
        console.error('Error:', error)
        alert('Failed to get response. Please try again.')
      }
    } finally {
      setIsLoading(false)
      abortControllerRef.current = null
    }
  }

  const handleNewChat = () => {
    setCurrentConversation(null)
    setMessages([])
    setStreamingMessage('')
  }

  const handleStopGeneration = () => {
    if (abortControllerRef.current) {
      abortControllerRef.current.abort()
    }
  }

  const handleSignOut = async () => {
    await supabase.auth.signOut()
  }

  return (
    <div className="chat-container">
      <div className="sidebar">
        <button onClick={handleNewChat} className="new-chat-btn">
          + New Chat
        </button>
        
        <div className="conversations-list">
          {conversations.map(conv => (
            <div
              key={conv.id}
              className={`conversation-item ${conv.id === currentConversation ? 'active' : ''}`}
              onClick={() => setCurrentConversation(conv.id)}
            >
              {conv.title}
            </div>
          ))}
        </div>

        <div className="user-info">
          <p>{user.email}</p>
          <button onClick={handleSignOut}>Sign Out</button>
        </div>
      </div>

      <div className="chat-main">
        <div className="messages-container">
          {messages.map((message, index) => (
            <Message key={index} {...message} />
          ))}
          
          {streamingMessage && (
            <Message role="assistant" content={streamingMessage} />
          )}
          
          <div ref={messagesEndRef} />
        </div>

        <form onSubmit={handleSubmit} className="input-form">
          <input
            type="text"
            value={input}
            onChange={(e) => setInput(e.target.value)}
            placeholder="Type your message..."
            disabled={isLoading}
          />
          {isLoading ? (
            <button type="button" onClick={handleStopGeneration}>
              Stop
            </button>
          ) : (
            <button type="submit" disabled={!input.trim()}>
              Send
            </button>
          )}
        </form>
      </div>
    </div>
  )
}

Step 4: Update App.tsx

Replace src/App.tsx:

import { useState, useEffect } from 'react'
import { supabase } from './lib/supabase'
import { User } from '@supabase/supabase-js'
import Auth from './components/Auth'
import Chat from './components/Chat'
import './App.css'

function App() {
  const [user, setUser] = useState<User | null>(null)
  const [loading, setLoading] = useState(true)

  useEffect(() => {
    // Check active sessions and sets the user
    supabase.auth.getSession().then(({ data: { session } }) => {
      setUser(session?.user ?? null)
      setLoading(false)
    })

    // Listen for changes on auth state
    const { data: { subscription } } = supabase.auth.onAuthStateChange((_event, session) => {
      setUser(session?.user ?? null)
    })

    return () => subscription.unsubscribe()
  }, [])

  if (loading) {
    return <div className="loading">Loading...</div>
  }

  return (
    <div className="app">
      {user ? <Chat user={user} /> : <Auth />}
    </div>
  )
}

export default App

Step 5: Add Styles

Update src/App.css:

* {
  margin: 0;
  padding: 0;
  box-sizing: border-box;
}

body {
  font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto', 'Oxygen',
    'Ubuntu', 'Cantarell', 'Fira Sans', 'Droid Sans', 'Helvetica Neue',
    sans-serif;
  background-color: #f7f7f8;
}

.app {
  height: 100vh;
  display: flex;
  flex-direction: column;
}

.loading {
  display: flex;
  justify-content: center;
  align-items: center;
  height: 100vh;
  font-size: 1.2rem;
  color: #666;
}

/* Auth Styles */
.auth-container {
  max-width: 400px;
  margin: 100px auto;
  padding: 2rem;
  background: white;
  border-radius: 12px;
  box-shadow: 0 2px 10px rgba(0, 0, 0, 0.1);
}

.auth-container h1 {
  margin-bottom: 0.5rem;
  color: #333;
}

.auth-container p {
  margin-bottom: 2rem;
  color: #666;
}

/* Chat Container */
.chat-container {
  display: flex;
  height: 100vh;
}

/* Sidebar */
.sidebar {
  width: 260px;
  background-color: #202123;
  color: white;
  display: flex;
  flex-direction: column;
  padding: 0.5rem;
}

.new-chat-btn {
  padding: 0.75rem;
  margin: 0.5rem;
  background-color: transparent;
  border: 1px solid #565869;
  border-radius: 6px;
  color: white;
  cursor: pointer;
  transition: background-color 0.2s;
}

.new-chat-btn:hover {
  background-color: #2a2b32;
}

.conversations-list {
  flex: 1;
  overflow-y: auto;
  margin: 1rem 0;
}

.conversation-item {
  padding: 0.75rem;
  margin: 0.25rem 0.5rem;
  border-radius: 6px;
  cursor: pointer;
  white-space: nowrap;
  overflow: hidden;
  text-overflow: ellipsis;
  transition: background-color 0.2s;
}

.conversation-item:hover {
  background-color: #2a2b32;
}

.conversation-item.active {
  background-color: #343541;
}

.user-info {
  padding: 1rem;
  border-top: 1px solid #565869;
}

.user-info p {
  margin-bottom: 0.5rem;
  font-size: 0.9rem;
}

.user-info button {
  width: 100%;
  padding: 0.5rem;
  background-color: transparent;
  border: 1px solid #565869;
  border-radius: 4px;
  color: white;
  cursor: pointer;
  transition: background-color 0.2s;
}

.user-info button:hover {
  background-color: #2a2b32;
}

/* Chat Main */
.chat-main {
  flex: 1;
  display: flex;
  flex-direction: column;
  background-color: #343541;
}

/* Messages Container */
.messages-container {
  flex: 1;
  overflow-y: auto;
  padding: 2rem 0;
}

/* Message Styles */
.message {
  padding: 1.5rem 0;
  border-bottom: 1px solid #4e4f60;
}

.message.user {
  background-color: #343541;
}

.message.assistant {
  background-color: #444654;
}

.message-role {
  max-width: 768px;
  margin: 0 auto 0.5rem;
  padding: 0 1rem;
  font-weight: 600;
  color: #d1d5db;
  text-transform: capitalize;
}

.message-content {
  max-width: 768px;
  margin: 0 auto;
  padding: 0 1rem;
  color: #d1d5db;
  line-height: 1.6;
}

/* Markdown Styles */
.message-content h1,
.message-content h2,
.message-content h3 {
  margin-top: 1.5rem;
  margin-bottom: 1rem;
}

.message-content p {
  margin-bottom: 1rem;
}

.message-content ul,
.message-content ol {
  margin-bottom: 1rem;
  padding-left: 2rem;
}

.message-content code {
  background-color: #2d2d30;
  padding: 0.2rem 0.4rem;
  border-radius: 3px;
  font-size: 0.9em;
}

.message-content pre {
  margin: 1rem 0;
  border-radius: 6px;
  overflow-x: auto;
}

.message-content pre > div {
  padding: 1rem !important;
}

/* Input Form */
.input-form {
  max-width: 768px;
  width: 100%;
  margin: 0 auto;
  padding: 1rem;
  display: flex;
  gap: 0.75rem;
}

.input-form input {
  flex: 1;
  padding: 0.75rem 1rem;
  background-color: #40414f;
  border: 1px solid #565869;
  border-radius: 6px;
  color: white;
  font-size: 1rem;
  outline: none;
  transition: border-color 0.2s;
}

.input-form input:focus {
  border-color: #10a37f;
}

.input-form input::placeholder {
  color: #8e8ea0;
}

.input-form button {
  padding: 0.75rem 1.5rem;
  background-color: #10a37f;
  border: none;
  border-radius: 6px;
  color: white;
  font-weight: 500;
  cursor: pointer;
  transition: background-color 0.2s;
}

.input-form button:hover:not(:disabled) {
  background-color: #1a7f64;
}

.input-form button:disabled {
  opacity: 0.5;
  cursor: not-allowed;
}

/* Mobile Responsiveness */
@media (max-width: 768px) {
  .chat-container {
    flex-direction: column;
  }

  .sidebar {
    width: 100%;
    height: auto;
    order: 2;
  }

  .chat-main {
    order: 1;
  }

  .conversations-list {
    max-height: 200px;
  }
}

Implementing Streaming Responses

One of the key features that makes our ChatGPT clone feel responsive is streaming responses. Let's dive deeper into how this works:

How SSE (Server-Sent Events) Work

Our Edge Function returns a ReadableStream with Content-Type: text/event-stream. This allows us to send chunks of data as they arrive from OpenAI, rather than waiting for the complete response.

Key benefits:

Lower perceived latency
Better user experience
Ability to cancel long-running requests
Real-time feedback

Client-Side Stream Processing

The React component processes the stream using the Fetch API's ReadableStream:

const reader = response.body?.getReader()
const decoder = new TextDecoder()

while (true) {
  const { done, value } = await reader.read()
  if (done) break
  
  // Process chunks as they arrive
  const chunk = decoder.decode(value)
  // Parse and display the content
}

Adding Conversation History

Our implementation includes several features for managing conversation history:

Auto-Titling

When creating a new conversation, we automatically generate a title from the first message:

title: messages[0]?.content.substring(0, 50) + '...' || 'New Chat'

Conversation Switching

Users can seamlessly switch between conversations, with messages loading dynamically from the database.

Persistence

All messages are stored in Supabase, ensuring conversations persist across sessions and devices.

Performance Optimization

1. Database Indexing

We've added indexes on foreign keys for faster queries:

CREATE INDEX idx_conversations_user_id ON conversations(user_id);
CREATE INDEX idx_messages_conversation_id ON messages(conversation_id);

2. Efficient Message Loading

Messages are loaded only when switching conversations, reducing unnecessary database calls.

3. Stream Buffering

We process stream chunks efficiently, updating the UI only when new content arrives.

4. Connection Pooling

Supabase Edge Functions automatically handle connection pooling to the database.

Security Best Practices

1. Row Level Security (RLS)

All tables have RLS enabled with policies ensuring users can only access their own data:

CREATE POLICY "Users can view own conversations" ON conversations
  FOR SELECT USING (auth.uid() = user_id);

2. API Key Protection

The OpenAI API key is stored securely in Edge Function secrets, never exposed to the client.

3. Authentication Required

All API calls require a valid Supabase auth token:

const {
  data: { user },
} = await supabaseClient.auth.getUser()

if (!user) {
  return new Response('Unauthorized', { status: 401 })
}

4. Rate Limiting

While not implemented in this tutorial, you can add rate limiting using the usage_stats table:

// Check user's recent usage
const { data: recentUsage } = await supabaseClient
  .from('usage_stats')
  .select('tokens_used')
  .eq('user_id', user.id)
  .gte('created_at', new Date(Date.now() - 3600000).toISOString()) // Last hour

const totalTokens = recentUsage?.reduce((sum, stat) => sum + stat.tokens_used, 0) || 0

if (totalTokens > 10000) { // 10k tokens per hour limit
  return new Response('Rate limit exceeded', { status: 429 })
}

Deployment

Step 1: Build the React App

npm run build

Step 2: Deploy to Vercel

npm install -g vercel
vercel

Follow the prompts to deploy your app. Make sure to add your environment variables in the Vercel dashboard.

Step 3: Update CORS Headers

Update your Edge Function to allow requests from your production domain:

const corsHeaders = {
  'Access-Control-Allow-Origin': process.env.NODE_ENV === 'production' 
    ? 'https://your-domain.vercel.app' 
    : '*',
  'Access-Control-Allow-Headers': 'authorization, x-client-info, apikey, content-type',
}

Step 4: Monitor Usage

Use the Supabase dashboard to monitor:

Edge Function invocations
Database usage
Auth metrics
Error logs

Advanced Features to Add

1. Model Selection

Add a dropdown to switch between GPT-3.5 and GPT-4:

<select value={model} onChange={(e) => setModel(e.target.value)}>
  <option value="gpt-3.5-turbo">GPT-3.5 Turbo</option>
  <option value="gpt-4">GPT-4</option>
</select>

2. System Prompts

Allow users to set custom system prompts for different conversation styles.

3. Export Conversations

Add functionality to export conversations as Markdown or PDF.

4. Search

Implement full-text search across all conversations:

-- Add search function
CREATE OR REPLACE FUNCTION search_messages(search_query TEXT, user_uuid UUID)
RETURNS TABLE (
  conversation_id UUID,
  message_content TEXT,
  message_role TEXT,
  created_at TIMESTAMPTZ
) AS $
BEGIN
  RETURN QUERY
  SELECT m.conversation_id, m.content, m.role, m.created_at
  FROM messages m
  JOIN conversations c ON m.conversation_id = c.id
  WHERE c.user_id = user_uuid
  AND m.content ILIKE '%' || search_query || '%'
  ORDER BY m.created_at DESC;
END;
$ LANGUAGE plpgsql;

5. Voice Input

Integrate the Web Speech API for voice input:

const recognition = new webkitSpeechRecognition()
recognition.onresult = (event) => {
  const transcript = event.results[0][0].transcript
  setInput(transcript)
}

Troubleshooting Common Issues

CORS Errors

Ensure your Edge Function includes proper CORS headers and handles OPTIONS requests.

Authentication Errors

Check that:

Your Supabase URL and anon key are correct
RLS policies are properly configured
The user is authenticated before making requests

Streaming Not Working

Verify that:

Your Edge Function returns the correct Content-Type
The client properly processes the stream
No proxy or CDN is buffering the response

Database Connection Issues

Check your RLS policies
Ensure indexes are created
Monitor the Supabase dashboard for errors

Cost Considerations

Supabase Costs

Free tier: 500MB database, 2GB bandwidth, 50k Edge Function invocations
Pro tier ($25/month): 8GB database, 250GB bandwidth, 2M Edge Function invocations

OpenAI Costs

GPT-3.5 Turbo: $0.002 per 1K tokens
GPT-4: $0.03 per 1K tokens (input), $0.06 per 1K tokens (output)

Optimization Tips

Implement token limits per message
Cache common responses
Use GPT-3.5 for most queries
Implement user quotas

Conclusion

Congratulations! You've built a fully functional ChatGPT clone with:

Real-time streaming responses
Persistent conversation history
User authentication
Responsive design
Secure API handling

This implementation demonstrates the power of Supabase Edge Functions for building modern AI applications. The serverless architecture ensures your app scales automatically while keeping costs low.

Next Steps

Add the advanced features mentioned above
Implement proper error handling and retry logic
Add analytics to track usage patterns
Consider implementing a credit system for premium features
Optimize for mobile with a Progressive Web App

Resources

Happy coding! 🚀

Table of Contents