ChatKey is small tool that enables you to use your own ChatGPT/GPT-4 prompts in any application that supports text input.
- Please ensure that you have configured the OPENAI_TOKEN environment variable with your API key
- Download the latest release
- Extract all files from the zip
- Run
ChatKey.exe - Start typing in any application that supports text input
- Select the text to use as input for the prompt
- Press the hotkey to show the popup menu (default:
Alt + .). - Select the prompt from the popup menu
- Wait for the response to be generated
- Review the generated response and press
Enter
I versus AI has done an incredible video covering most topics:
To configure ChatKey, you can edit the config.ini file provided.
| Key | Description | Default |
|---|---|---|
popup_menu_hotkey |
The hotkey to show the popup menu | Alt + . |
replace_text |
Whether to replace the selected text with the response | 0 |
show_notification |
Whether to show a notification when generating a response | 1 |
max_input_length |
The maximum length of input text (0 = unlimited) | 0 |
To add new prompts, you must include a new section in the config.ini file. For instance, if you wish to include a prompt for translating text to French, you can achieve this by appending the following section to the configuration file:
[prompt_translate_to_french]
name = Translate to French
shortcut = t
system_prompt = "I want you to act as a French translator. I will say something in any language and you will translate it to French. The first thing I want you to translate is:"
temperature = 0.2
model = gpt-3.5-turboTo ensure that the newly added prompt is available in the popup menu, it must be included in the [popup_menu] section. Additionally, if you have already configured multiple prompts, you can tidy up the popup menu by utilizing --- as a separator.
[popup_menu]
---
prompt_translate_to_frenchThe changes will be applied automatically, there's no need to restart ChatKey (the only exception to this rule is the global popup_menu_hotkey).
You can individually configure the parameters of each prompt. If keys with default values are omitted, the default values will be used instead.
| Key | Description | Default |
|---|---|---|
name |
The name of the prompt that will be displayed in the popup menu | |
shortcut |
The shortcut key to select the prompt from the popup menu | |
system_content |
The prompt that will be used to generate the response (required) | |
model |
The model to use when generating the response, more info here | gpt-3.5 |
temperature |
The temperature to use when generating the response (0.0 - 2.0) | 0.7 |
top_p |
The top_p to use when generating the response (0.0 - 1.0) | |
presence_penalty |
Increase the model's likelihood to talk about new topics (-2.0 - 2.0) | |
frequency_penalty |
Decrease the model's likelihood to repeat the same line verbatim (-2.0 - 2.0) |
- I versus AI for the awesome tutorial video
- cocobelgica for the JSON lib
- teadrinker for the HBitmapFromResource lib
The code in this repository is licensed under the MIT License. See LICENSE for more information.