Skip to content

An embeddable widget for interacting with openAI api compatable LLM's

License

Notifications You must be signed in to change notification settings

helixml/chat-widget

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

41 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

@helixml/chat-widget

helix logo

The @helixml/chat-widget is a UI component that makes it easy to expose large language models behind common APIs (e.g. OpenAI API). It is packaged both as a react component and a browser-side library.

screenshot

Browser Usage

The @helixml/chat-embed library is a standalone wrapper around our React library that can easily be embedded onto any page with a <script> tag.

It will bundle using the IIFE format which means that we can put this script tag on any page that uses any JS library - it is totally self contained and non combative with other versions of React or other libraries the user might have installed.

Install

Load the library from a CDN and then you can call the global ChatWidget to render it somewhere in your application.

You can find the full list of versions on jsdelivr. You can use https://www.srihash.org/ to calculate the integrity hash.

The arguments you pass to the ChatWidget function are the same as the properties passed to the React library.

<script src="https://cdn.jsdelivr.net/npm/@helixml/chat-embed@VERSION" integrity="..." crossorigin="anonymous"></script>
<script>
  ChatWidget({
      url: 'https://api.openai.com/v1/chat/completions',
      model: 'gpt-3.5-turbo',
      bearerToken: 'YOUR_API_KEY',
  })
</script>

Examples

You can see some plain html examples in the embed/examples directory.

React Component Usage

The @helixml/chat-widget is a highly customizable React component designed to provide interactive chat functionality within your application. It features a minimalist design that opens a modal window upon interaction, where users can submit questions and receive answers from a specified openAI compatible endpoint.

Installation

To use the @helixml/chat-widget in your project, install it via npm:

npm install @helixml/chat-widget

or using yarn:

yarn add @helixml/chat-widget

Usage

Import and use the Widget component in your React application:

import React from 'react';
import Widget from '@helixml/chat-widget';

function App() {
  return (
    <div className="App">
      <Widget
        url="https://myopenaiendpoint.com"
        model="my_model_name"
        windowTheme={{ /* Optional window theme overrides */ }}
        searchTheme={{ /* Optional search theme overrides */ }}
      />
    </div>
  );
}

export default App;

Props

  • url: String. The endpoint URL for the remote API from which the answers will be fetched.
  • model: String. Identifier for the specific model to be queried at the remote API.
  • bearerToken: (Optional) String. Bearer token for authentication with the remote API, if required.
  • title: (Optional) String. Title text for the chat window.
  • logo: (Optional) String. URL of the logo image to display in the header.
  • placeholder: (Optional) String. Placeholder text for the search input.
  • windowTheme: (Optional) Object. Theme customization for the window component. See the Theme Customization section below for details.
  • searchTheme: (Optional) Object. Theme customization for the search input. See the Theme Customization section below for details.

Theme Customization

Customize the appearance of the chat widget and its components using the windowTheme and searchTheme props. Each theme object allows you to override default styles to match your application's design.

Window Theme Options

  • logoWidth: Width of the logo inside the header.
  • backdropColor: Color of the modal backdrop.
  • backgroundColor: Background color of the window.
  • width: Width of the window.
  • verticalMargin: Vertical margin of the window.
  • borderRadius: Border radius of the window.
  • shadow: Box shadow for the window.
  • fontFamily: Font family for text inside the window.
  • headerTextColor: Text color for the header.
  • headerFontSize: Font size for the header text.
  • headerPadding: Padding inside the header.
  • errorTextColor: Text color for error messages.
  • errorFontSize: Font size for error messages.
  • contentBoxShadow: Box shadow for the content area inside the window.
  • contentPadding: Padding inside the content area.
  • replyTextColor: Text color for the reply messages.
  • replyFontSize: Font size for the reply messages.
  • subtitleTextColor: Text color for subtitles.
  • subtitleFontSize: Font size for subtitles.
  • footerPadding: Padding inside the footer.
  • closeButtonBorderRadius: Border radius of the close button.
  • closeButtonColor: Color of the close button.
  • closeButtonPadding: Padding of the close button.
  • closeButtonFontSize: Font size of the close button text.

Search Theme Options

  • borderColor: Border color of the search input.
  • backgroundColor: Background color of the search input.
  • hoverBorderColor: Border color of the search input on hover.
  • borderRadius: Border radius of the search input.
  • iconPadding: Padding around the search icon.
  • iconColor: Color of the search icon.
  • textPadding: Padding inside the search input for text.
  • textSize: Font size of the text inside the search input.
  • fontFamily: Font family of the search input text.

License

This project is licensed under the MIT License - see the LICENSE file for details.

About

An embeddable widget for interacting with openAI api compatable LLM's

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •