The recently announced GPT-4 Chat model will work the same way as the model demonstrated in this video, except the token limit has increased substantially.

If you find my content useful and want to support the channel, consider contributing a coffee ☕: https://hbyt.us/coffee

In this video, we cover how to build and deploy your own full-stack AI application using SvelteKit, OpenAI's newest ChatGPT API / Chat Completion Model (gpt-3.5-turbo) APIs, and Vercel Edge Functions. The new ChatGPT API is 10x cheaper and much faster than GPT-3. We also learn how to handle moderations, token limits, and more.

📁 Starting Code: https://github.com/huntabyte/chatty/t...
📁 Final Source Code: https://github.com/huntabyte/chatty

OpenAI Playground: https://platform.openai.com/playgroun...
OpenAI Tokenizer: https://platform.openai.com/tokenizer
GPT-3 Tokenizer: https://npmjs.com/gpt3-tokenizer
SSE.js: https://www.npmjs.com/package/sse.js
SvelteKit Config Docs: https://kit.svelte.dev/docs/adapter-v...

🚀 Become a channel member: https://hbyt.us/join
💬 Discord: https://hbyt.us/discord
🐦 Twitter: https://hbyt.us/twitter
🖥️ Setup Stuff: https://hbyt.us/gear

Chapters:
00:00 - What we're building
00:20 - How the model works
01:40 - Starting Code
02:08 - Chat Endpoint/API Route
04:55 - Setup Tokenizer
06:24 - Finish Endpoint/API Route
12:20 - Install & Setup SSE
12:49 - Client-Side Route
21:05 - Deployment Configs
21:54 - Deploy to Vercel

📃 Topics Covered:
- SvelteKit Streaming
- SvelteKit AI App
- SvelteKit Vercel Edge Functions
- Vercel Edge Functions
- Vercel AI App
- ChatGPT API
- GPT4 API
- GPT4 Web App
- GPT 4 ChatBot
- OpenAI Node App
- OpenAI WebApp
- OpenAI ChatGPT
- Custom Chatbot
- ChatGPT API
- Custom ChatGPT
- gpt-3.5-turbo model
- ChatGPT OpenAI App
- ChatGPT App
- OpenAI Completions
- OpenAI Chat Completions
- OpenAI Moderations