AI-Powered Extensions for VCL
- 4 minutes to read
AI-powered Extensions leverage advanced natural language processing (NLP) technologies to offer automated, intelligent text editing functionality directly within your VCL applications.
Note
All DevExpress AI-powered Extensions follow the bring your own key principle. DevExpress does not offer a REST API and does not ship any built-in LLMs/SLMs. You need an active Azure, OpenAI, or other AI service subscription to obtain the REST API endpoint, key, and model deployment name.
AI-Powered Commands for Native VCL Controls
All DevExpress controls listed in this section can display a context menu populated with a list of predefined AI-powered user commands (AI Assistant) if an AI service provider is registered and AI user commands are initialized.
Predefined AI-Powered Commands
- Change Style
- Rephrases or paraphrases selected text while retaining its original meaning (12 predefined styles available).
- Change Tone
- Adjusts the text tone to meet specific audience or context requirements (Casual, Confident, Friendly, Professional, Straightforward).
- Expand
- Expands original text with additional information or in-depth explanations.
- Explain
- Rephrases text in more clear terms and makes complex content more accessible and understandable.
- Proofread
- Reviews text for spelling, grammar, punctuation, and style errors.
- Shorten
- Removes redundant details and makes content more concise.
- Summarize
- Creates a brief summary based on original text.
- Translate
- Translates original text into a different language (English, German, French).
Native VCL Controls with Support for AI Assistant Commands
ExpressRichEditControl
- TdxRichEditControl
A fully-functional rich text editor with advanced formatting and display capabilities you expect from a modern word processor. If an AI service provider is registered and AI user commands are initialized, the Rich Edit control can display these commands grouped within a context menu’s AI Assistant item.
You can handle the TdxRichEditControl.OnGetAICommands event to manage the list of available AI-powered commands and change their status depending on specific conditions in your application. For example, you can hide or disable AI Assistant commands if no content is selected in the editor.
Tip
Refer to the TdxRichEditControl.OnGetAICommands event description for additional information and code examples.
Multi-Line Text Editors
- TcxRichEdit
- A simple rich text editor (a DevExpress wrapper for the standard TRichEdit component; adds support for skins and deep integration with other DevExpress VCL components).
- TcxMemo | TcxDBMemo
- Unbound and data-aware memo editors for plain text content (multi-line text boxes).
Similar to TdxRichEditControl, all listed ExpressEditors can display the same list of AI-powered commands if an AI service is registered; you can also handle the Properties.OnGetAICommands event for any supported editor to manage the list of available AI Assistant commands.
AI-Powered Command Implementation for Native VCL Controls
In v25.1, we ship no dedicated APIs for native DevExpress VCL controls. Until Embarcadero ships official AI-related SDK libraries, VCL developers can plug in third-party libraries or leverage their own implementation to support different AI providers.
GitHub Example: Third-Party Library-Based Solution
We published a GitHub example that demonstrates a custom TdxAIChatClient implementation for a third-party AI library for VCL (GenAI). You can download it using the GetIt Package Manager.
Related Compiled Demo
To see AI-powered commands in action, run the Word Processing RTF demo in the VCL Demo Center installed with compiled VCL DevExpress demos.
The Word Processing RTF demo uses our Azure OpenAI Service deployment implemented as the TdxAIAzureChatClient class. This class is declared in the dxAI.ChatClient.Azure.pas unit shipped with demo source code.
Tip
You can find full source code for the installed Rich Edit control demo in the following folder:
%PUBLIC%\Documents\DevExpress VCL Demos\MegaDemos\Product Demos\ExpressRichEditControl
DevExpress AI-Powered Extensions for ExpressReports
The TdxReport component supports the same AI-powered extensions available for DevExpress Web Reports – Summarize, Translate, Preview Generation, etc.
Refer to the following topic for detailed information in this regard: AI-Powered Extensions for DevExpress Reporting.
Enable AI-Powered Extensions for ExpressReports
You can enable DevExpress AI-powered Extensions in your ExpressReports-based application in the Project Settings dialog. Open the Reports tab and specify REST API endpoint URI, valid API key, and model/deployment name for the selected AI service.
Related Compiled Demo
To see the TdxReport component in action (including AI-powered capabilities), run the Report Designer/Viewer demo in the VCL Demo Center installed with compiled VCL DevExpress demos. Click different items in the sidebar on the left to switch between demo features.
Tip
You can find full source code for the installed compiled Report demo in the following folder:
%PUBLIC%\Documents\DevExpress VCL Demos\MegaDemos\Product Demos\ExpressReports