Skip to content

README.md

Latest commit

 

History

History
34 lines (23 loc) · 948 Bytes

README.md

File metadata and controls

34 lines (23 loc) · 948 Bytes

OpenAI Stream Example

This code uses the Fetch API to return a streaming response from the OpenAI API.

NOTE: The Completion API will be deprecated starting January 2024. Please use the ChatCompletion API instead, which can be found in models/Chat.js

NOTE 2: GPT-4 is only available to current users of the OpenAI API. If you are new, it will not work! In that case, ise gpt-3.5-turbo instead of gpt-4 as your model

How to use

Instantiate a class in the following way:

const config = {
  model: 'gpt-4',
  purpose: "You are a snarky teen who will only answer questions if the user gives you something."
}

const callback = chunk => response += chunk

const chat = new Chat(config, callback)
chat.setApiKey = "YOUR API KEY"

See the example directory to see how it could be used in an app.
See the models directory to see how streaming functions

Run example

$ cd example
$ npm install
$ npm run dev