Mistral AI ChatBox

Welcome to the Mistral AI ChatBox! This application leverages the power of Mistral AI to provide intelligent, context-aware chat capabilities. Designed for developers and enthusiasts alike, the Mistral AI ChatBox is your go-to solution for building interactive, AI-driven chat applications.

Description

The Mistral AI ChatBox is a feature-rich template that allows you to create cross-platform desktop applications with advanced AI capabilities. By combining Electron for desktop functionality and Next.js for modern web development, this app provides a seamless development experience. With built-in support for Markdown and LaTeX parsing, as well as syntax-highlighted code responses, the Mistral AI ChatBox is perfect for developers looking to build sophisticated, AI-powered chat interfaces.

Note: A valid Mistral AI API key is required to use this application. Be sure to configure your API key in the config.mjs file before running the app.

Configuration

Before running the application, ensure you configure the required values in the config.mjs file. These settings are essential for connecting to the Mistral AI API and customizing its behavior.

Steps to Configure

1. Locate the config.mjs file in the project directory.
2. Replace the placeholder api_key value with your actual Mistral AI API key.
3. Adjust the completionOptions settings as needed to suit your application's requirements.

Mistral AI Defaults

/** *************************************************************************
 * Mistral AI defaults
 */
Mistral: {
  /**
   * The API key for the Mistral AI.
   */
  api_key: 'your-api-key-here', // Replace with your actual API key
  /**
   * The completion body options.
   */
  completionOptions: {
    /**
     * The model to use for the Mistral AI.
     */
    model: 'mistral-large-latest', // Options: 'mistral-large-latest', 'mistral-small-latest', 'codestral-latest'
    /**
     * The top-p value for sampling. Controls the diversity of the generated output.
     */
    top_p: 0.8,
    /**
     * The maximum number of tokens to generate.
     */
    max_tokens: 4096,
    /**
     * Whether to stream the output or return it as a single response.
     */
    stream: true,
    /**
     * Whether to use safe prompt mode. When enabled, the AI will avoid generating harmful or unsafe content.
     */
    safe_prompt: true,
    /**
     * The random seed to use for generating the output.
     */
    // random_seed: 1337
  }
}

Development Mode Configuration

To ensure the app runs correctly in development mode, you need to set the isDev boolean inside main.js. This flag determines whether the app is running in development or production mode. In the future, this configuration will be moved to a dedicated config/env file for better maintainability.

Once configured, you can proceed with running or building the application as described in the Installation and Build Process sections.

Note: Ensure you keep your API key secure and do not expose it in public repositories.

Installation

Getting started is easy! Follow these steps to set up your development environment:

1. Clone the repository:

   git clone https://github.com/gidsola/Mistral-AI-ChatBox.git
   cd electron-nextjs-starter-app

2. Install dependencies:

   npm install

3. Open two bash terminals:

   • In the first terminal, run the Next.js development server:

     npm run dev

   • In the second terminal, start the Electron app:

     npm start

And that's it! You should now have the app up and running.

Build Process

Building the application is straightforward. Here are the commands you need:

• Build Next.js:

  npm run build:next

• Build Electron:

  npm run build:electron

• Build both Next.js and Electron:

  npm run build

These commands will compile your application and prepare it for distribution.

Preview

Check out a quick preview of the app in action:

Happy coding! 🚀