This is a Windows tool that enables running custom AI prompts on text in any window using global hotkeys.
i.e. Low-friction AI text editing ("spicy autocomplete") anywhere in Windows.
Where can it be used?
Almost anywhere in Windows where you can enter text.
To get started, first download and extract the latest release .zip file. If you already have AutoHotkey installed, simply run AI-Tools.ahk. If not, use the .exe version, which allows you to use the script without having AutoHotkey installed. The script doesn't install anything and is portable, so you can run it from any location.
When you run the script for the first time, it will create a new settings.ini file in the same directory. This file contains the script's settings, which you can edit to change the hotkeys or add your own prompts.
Additionally, the script will prompt you to enter your API key. You can obtain an API key from:
- OpenAI: https://platform.openai.com/
- Google Gemini: https://aistudio.google.com/app/apikey
The default hotkeys and prompts are set to the following:
Ctrl+Shift+j - (Auto-select text - Fix spelling) - Auto selects the current line or paragraph and runs the "Fix Spelling" prompt and replaces it with the corrected version.
Ctrl+Shift+k - (Auto-select text - Prompt Menu) - Auto selects the current line or paragraph and opens the prompt menu.
Ctrl+Alt+Shift+k - (Manual-select text - Prompt Menu) - Opens the prompt menu to pick the prompt to run on the selected text.
The settings.ini file contains the settings for the script. You can edit this file to change the prompts, the API mode and model to use, and individual model settings.
Start with windows
To have the script start when windows boots up, select "Start With Windows" from the tray icon.
This tool supports multiple AI API providers:
- Endpoint:
https://api.openai.com/v1/chat/completions - Models: gpt-4, gpt-4-turbo, gpt-3.5-turbo, etc.
- API Key: https://platform.openai.com/
- Endpoint:
https://[resource].openai.azure.com/openai/deployments/[model]/chat/completions - Models: Your deployed models
- Documentation: Azure OpenAI Quickstart
- Endpoint:
https://generativelanguage.googleapis.com/v1beta/models/{model}:generateContent - Models: gemini-2.5-flash, gemini-2.5-flash-lite, etc.
- API Key: https://aistudio.google.com/app/apikey
- Documentation: Gemini API Docs
To switch providers, edit the default_mode setting in settings.ini:
[settings]
default_mode = mode_chat_completion ; OpenAI (default)
; default_mode = mode_chat_completion_azure ; Azure OpenAI
; default_mode = mode_gemini ; Google GeminiIndividual prompts can override the mode by setting mode=mode_gemini in their prompt section.
Tested on:
- Windows 10 Pro 22H2 64-bit
- Windows 11 Pro 25H2
TheArkive (JXON_ahk2, M-ArkDown_ahk2), iseahound (SetSystemCursor), and the AHK community.
