A rust Chat GPT-3 CLI tool.

This implementation of the main function calls the appropriate subcommand based on the user's input, using the provided auth_token and model values to create a Client instance for sending requests to the OpenAI API.

When the chat subcommand is called, the CompletionResponse struct is used to deserialize the JSON response from the OpenAI API, and the generated chat response is printed to the console.

When the code subcommand is called, the CodeCompletionResponse struct is used to deserialize the JSON response from the OpenAI API, and the generated code completion is printed to the console.

When the image subcommand is called, a Vec<String> of URLs for the generated images is returned from the OpenAI API, and each URL is printed to the console using a for loop.

Overall, this implementation provides a full-featured CLI tool for working with OpenAI's GPT-3 API in Rust.