As someone who didn't want to pay exorbitant prices for api access, I decided to use the magic of JavaScript to inject my questions directly into the HTML bodies of the LLM webpages. The app leverages Electron's ipc API to pass messages between the app context itself and the browser instances that are loaded, allowing for real-time prompting when a key is pressed down.
The result is a fully functional desktop app that allows you to simultaneously prompt multiple LLMs at the same time through their browser interfaces!
Right now, the app supports multiple LLM's like
- ChatGPT - Gemini - Claude - Grok - DeepSeek - Copilot
In addition, you can save a frequently asked question or prompt in the app, and easily access it later, all in one place. This saves time and energy!
- Demo video is here: https://www.youtube.com/watch?v=YxqWUp0Wmi0
- Code Walkthrough is here: https://www.youtube.com/watch?v=bkSRSUMsh10
The project is open source and runs locally. I would love any feedback or new features to possibly add!
Thank you for reading!