AI Local

This is a localhost implementation of the Open AI API. This will allow the user to interface with the API locally without having to host the file in the cloud. This is not a chat interface rather a quick option to interface with the API.

Another benefit of using APIs is that the prompts/responses are not used to train the models (see below)

API	Used to train models
OpenAI	No (as per privacy docs)
Gemini	Yes
Claude	Unknown

Features

• Supports creating images using DALL-E 3
  • shows the revised prompt that OpenAI API generates
• Allows the user to monitor daily cost for OpenAI API usage
• Markdown support in responses
• Supports <code> blocks
  • language specific highlighting
  • copy button to copy code
• Supports dark mode

Usage

• Edit the script.js file
  • Add your OpenAI API key in const OPENAI_API_KEY
  • Add your Google Gemini API Key in const GEMINI_API_KEY

OpenAI

Currently supported API endpoints are:

• GPT-4 Omni
• GPT-3.5 Turbo
• DALL-E 3
• GPT-4 Vision Preview
• TTS-1 (text-to-speech)

Planned features

• Add support for Gemini by Google
• Better error handling
• Add chat feature
• Add support for streaming
• Add web search support
• Allow more api parameters to be configured for each endpoint
• Add support for OpenAI speech-to-text
• Add support for Claude by Anthropic (currently has issue with CORS)
• Implement tokenizer to display token count for prompt
• Add cost calculator for Gemini and Claude