RAD Studio - Recent changes [en]

Delphi Compiler Directives (List) Index

Nicolas.rios — Thu, 10 Sep 2026 21:04:29 GMT

Compiler Directives

← Older revision		Revision as of 21:04, 10 September 2026
Line 8:		Line 8:

	==Compiler Directives==		==Compiler Directives==

			{{Note\|Local and Global describe the extent to which a compiler directive setting applies during compilation.
			* Local directives affect the current unit from the position they are declared to the end of the source code file. The settings can be changed later within the source file.
			* Global directives affect the entire compiled module or the entire project compilation.}}

	{\| border="1" frame="hsides" rules="rows"		{\| border="1" frame="hsides" rules="rows"
	! '''{{Delphi}} Compiler Directive'''		! '''{{Delphi}} Compiler Directive'''

What's New

Yukikoi — Wed, 09 Sep 2026 14:27:17 GMT

Replaced the <references> tag with {{#tag:references}} to prevent the tag from being exposed in localized pages after WikiManager processing. No change to the rendered output.

← Older revision		Revision as of 14:27, 9 September 2026
Line 633:		Line 633:

	=== References ===		=== References ===
	<references>		<!-- Keep one blank line between each <ref> entry. WikiManager uses the blank lines to export each reference as a separate XLF trans-unit. -->
			{{#tag:references\|

	<ref name=”TAniCalculations”> The reason for removing the animation render interval lies in the fact that it now consistently matches the current refresh rate of the display. The application can run on devices with 60 Hz, 90 Hz, or 120 Hz displays, and scrolling will automatically become smoother, without the need for manual interval adjustments. The animations that are running will constantly be updated (recalculated) on each new frame of the screen. </ref>		<ref name=”TAniCalculations”> The reason for removing the animation render interval lies in the fact that it now consistently matches the current refresh rate of the display. The application can run on devices with 60 Hz, 90 Hz, or 120 Hz displays, and scrolling will automatically become smoother, without the need for manual interval adjustments. The animations that are running will constantly be updated (recalculated) on each new frame of the screen. </ref>

	~~</references>~~		}}

	[[de:Neuerungen]]		[[de:Neuerungen]]
	[[fr:Nouveautés]]		[[fr:Nouveautés]]
	[[ja:新機能]]		[[ja:新機能]]

Configuring Ollama for Smart CodeInsight

L10nBot — Wed, 09 Sep 2026 14:23:27 GMT

updated language links

← Older revision		Revision as of 14:23, 9 September 2026
Line 72:		Line 72:
	</li>		</li>
	</ol>		</ol>

			[[de:Ollama für Smart CodeInsight konfigurieren]]
			[[ja:Smart CodeInsight 用 Ollama の設定]]

Installation Notes

Yukikoi — Wed, 09 Sep 2026 09:45:43 GMT

Optional Additional Software: Removed unnecessary line breaks that caused formatting issues in three localized versions.

Show changes

Supported Target Platforms

Yukikoi — Wed, 09 Sep 2026 07:03:52 GMT

Show changes

Smart CodeInsight

Yukikoi — Wed, 09 Sep 2026 02:58:46 GMT

Changed to wiki markup to avoid XLF output errors caused by incorrect <ul><li> markup.

← Older revision		Revision as of 02:58, 9 September 2026
Line 1:		Line 1:
	[[Smart CodeInsight]] is an open architecture that allows you to work with your favorite AI engines. {{Product}} also offers you customizable UI integration, besides the following implemented features:		[[Smart CodeInsight]] is an open architecture that allows you to work with your favorite AI engines. {{Product}} also offers you customizable UI integration, besides the following implemented features:
	~~<ul>~~
	~~<li>~~A core AI engine and ToolsAPI.~~</li>~~		* A core AI engine and ToolsAPI.
	~~<li>~~ IDE integration:		* IDE integration:
	~~<ul>~~		** [[#Configuration Options\|Configuration options for the engines]].
	~~<li>~~ [[#Configuration Options\|Configuration options for the engines]].~~</li>~~		** [[#Editor Menu Commands\|Editor menu commands to invoke AI operations]].
	~~<li>~~ [[#Editor Menu Commands\|Editor menu commands to invoke AI operations]].~~</li>~~		** [[#AI Chat Window\|AI chat window]].
	~~<li>~~ [[#AI Chat Window\|AI chat window]].~~</li>~~		* Support for online and offline LLM providers such as:
	~~</ul>~~		** OpenAI, Gemini, and Claude (online).
	~~</li>~~		** Ollama (offline).
	~~<li>~~Support for online and offline LLM providers such as:		::{{Tip\|Ollama is distributed as open source under an MIT license. Individual models that you might use with Ollama have their own licenses. You can find the Codellama Acceptable Use Policy on this page.}}
	~~<ul>~~		::{{Note\|For ToolsAPI enthusiasts, {{Product}} gives you the opportunity to create your own API plugin. For more information on this, please refer to [[Create and implement your AI plugin]] page.}}
	~~<li>~~ OpenAI, Gemini, and Claude (online).~~</li>~~
	~~<li>~~ Ollama (offline).
	{{Tip\| Ollama is distributed as open source under an MIT license. Individual models that you might use with Ollama have their own licenses. You can find the Codellama Acceptable Use Policy on this page.}}
	~~</li>~~
	{{Note\| For ToolsAPI enthusiasts, {{Product}} gives you the opportunity to create your own API plugin. For more information on this, please refer to [[Create and implement your AI plugin]] page.}}
	~~</ul></li>~~
	~~</ul>~~
	{{Warning\|{{Product}} does not provide access to any of those engines. Using these AI services requires you to agree to the terms of the specific vendor, receive an API key to enter into the {{Product}} configuration, and pay the AI company the due fees.		{{Warning\|{{Product}} does not provide access to any of those engines. Using these AI services requires you to agree to the terms of the specific vendor, receive an API key to enter into the {{Product}} configuration, and pay the AI company the due fees.
	The only exception to this rule is Ollama, which can be used offline without incurring a service fee and installed locally or on a server of your choice.}}		The only exception to this rule is Ollama, which can be used offline without incurring a service fee and installed locally or on a server of your choice.}}

File:Kai custom settings.png

Nicolas.rios — Tue, 08 Sep 2026 15:10:49 GMT

Nicolas.rios uploaded File:Kai custom settings.png

New page

Kai custom settings

File:Kai built-in.png

Nicolas.rios — Tue, 08 Sep 2026 15:10:08 GMT

Nicolas.rios uploaded File:Kai built-in.png

New page

Kai built-in

File:Kai chat crop.png

Nicolas.rios — Tue, 08 Sep 2026 15:09:06 GMT

Nicolas.rios uploaded File:Kai chat crop.png

New page

Kai chat options

File:Kai chat v1.1.png

Nicolas.rios — Tue, 08 Sep 2026 15:07:47 GMT

Nicolas.rios uploaded File:Kai chat v1.1.png

New page

Kai chat v1.1

Kai

Nicolas.rios — Tue, 08 Sep 2026 15:07:01 GMT

Sections

← Older revision		Revision as of 15:07, 8 September 2026
Line 12:		Line 12:
	* [[Using Local AI Models with Kai]]		* [[Using Local AI Models with Kai]]
	* [[Kai Options]]		* [[Kai Options]]
			** [[Kai Options: Completions]]
	** [[Kai Options: Chat]]		** [[Kai Options: Chat]]
	** [[Kai Options: ~~Completions~~]]		*** [[Kai Options: ACP Registry]]
	** [[Kai Options: MCP Servers]]		*** [[Kai Options: MCP Servers]]
			**** [[Kai Options: Built-in]]
			**** [[Kai Options: Custom]]

	[[Category:Using {{Product}}]]		[[Category:Using {{Product}}]]

Main Page

Nicolas.rios — Tue, 08 Sep 2026 15:06:22 GMT

Kai AI Agent

← Older revision		Revision as of 15:06, 8 September 2026
Line 53:		Line 53:
	* [[Kai Installation]]		* [[Kai Installation]]
	* [[Kai Release Notes]]		* [[Kai Release Notes]]
			* [[Kai Version 1.1]]
	* [[Kai Features]]		* [[Kai Features]]
	* [[Kai Options]]		* [[Kai Options]]
	* [[Using Local AI Models with Kai]]		* [[Using Local AI Models with Kai]]
	* [[Kai Licensing]]		* [[Kai Licensing]]

	== Reference ==		== Reference ==
	* [[Supported Target Platforms\|Target Platforms]]		* [[Supported Target Platforms\|Target Platforms]]

Kai

Nicolas.rios — Tue, 08 Sep 2026 15:05:16 GMT

← Older revision		Revision as of 15:05, 8 September 2026
Line 7:		Line 7:
	* [[Kai Installation]]		* [[Kai Installation]]
	* [[Kai Release Notes]]		* [[Kai Release Notes]]
			* [[Kai Version 1.1]]
	* [[Kai Features]]		* [[Kai Features]]
	* [[Kai Licensing]]		* [[Kai Licensing]]

File:Kai v1.1 RedesignUI.png

Nicolas.rios — Tue, 08 Sep 2026 15:04:36 GMT

Nicolas.rios uploaded File:Kai v1.1 RedesignUI.png

New page

Kai v1.1 RedesignUI

File:Kai ACP Registry.png

Nicolas.rios — Tue, 08 Sep 2026 15:03:56 GMT

Nicolas.rios uploaded File:Kai ACP Registry.png

New page

Kai ACP Registry

File:Kai install v1.0.2.png

Nicolas.rios — Tue, 08 Sep 2026 15:02:23 GMT

Nicolas.rios uploaded File:Kai install v1.0.2.png

New page

Kai install v1.0.2

Kai Version 1.1

Nicolas.rios — Tue, 08 Sep 2026 15:01:33 GMT

Created page with "{{Parent|Kai}} The Kai 1.1 release, {{Product}}’s agentic AI extension, is a significant architectural enhancement over the original 1.0 version. The 1.1 release changes how Kai interacts with the various AI agents and the local CLI engine. As a result, it has several configuration and integration incompatibilities. In some cases, it is necessary to remove previous configurations to avoid incompatibilities – not in Kai, but in newer versions of the third-party CLI..."

New page

{{Parent|Kai}}

The Kai 1.1 release, {{Product}}’s agentic AI extension, is a significant architectural enhancement over the original 1.0 version.

The 1.1 release changes how Kai interacts with the various AI agents and the local CLI engine. As a result, it has several configuration and integration incompatibilities. In some cases, it is necessary to remove previous configurations to avoid incompatibilities – not in Kai, but in newer versions of the third-party CLIs.

==Kai 1.1 Installation==
Starting with Kai version 1.1, installation has fewer dependencies, thanks to the ACP registry.

The installer now ships only GitHub Copilot completions support (separate from chat support, delivered via the ACP registry) and support for both chat and completions for local engines.

[[File:Kai install v1.0.2.png|border|link=|700px]]

==Kai 1.1 Features==
This section details the core features in Kai version 1.1.

===Kai UI Redesign===
The Kai UI redesign has been extensive, focusing on many details, icons, colors, and more.

Kai 1.1 introduces:
* Monochrome icons and glyphs.
* Better spacing.
* An editing chat area color is more consistent with the “editing” operation.
* A new toolbar for the chat window.
* Cleanup configuration pages.
* Checkboxes at the bottom of the chat window for boolean options.

Below is the UI of the Kai chat window, with the new toolbar, icons, and overall look and feel:

[[File:Kai v1.1 RedesignUI.png|link=|border|600px]]

===ACP Registry Support===
{{Menu|Tools > Options > Kai > Chat > [[Kai_Options: ACP Registry|ACP Registry]]}}

With the new ACP Registry support, Kai no longer includes pre-installed providers (except for local models), and users no longer need to install the Claude Agent ACP manually via NPM.

Kai now supports a much wider range of agents, which users can manually install from the ACP Registry configuration page. The ACP Registry currently includes 35+ agents: all the agents Kai supported previously remain available, and more well-known agents have been added, such as Cursor, Devin, Kilo Code, Junie, etc.

You can view the full list and general information about the ACP Registry here: [https://agentclientprotocol.com/get-started/registry ACP Registry - Agent Client Protocol].

[[File:Kai ACP Registry.png|link=|border|800px]]

{{Attention| Zed Industries and JetBrains maintain the registry, and Embarcadero has no control over or responsibility for the registry or the agents downloaded from it.}}

Select the agents you want to download, click the '''Install''' button, and use them in Kai immediately.

Other agents are only available via NPX (node.js) or UVX (PyPi Python); in this case, an extra confirmation is required before installing. To install those agents, the user must manually configure their computer to support the underlying technology. For example, you need to install Node.JS to have support for NPX agents. Kai does not automatically install Node.JS on your system. Once the core engine (like Node.JS) is available, clicking '''Install''' makes the corresponding package available, rather than downloading a package binary directly.

Each agent is available in the latest version; if an update is available, an '''Update''' button appears (as shown above). Once the agent is installed, it appears in the chat '''New session''' and '''Import sessions''' dropdowns. It also appears in the {{Menu|Tools > Options > [[Kai Options: Chat|Kai Chat]]}}.

===MCP Servers Update===
{{Menu|Tools > Options > Kai > Chat > [[Kai_Options: MCP Servers|MCP Servers]]}}

Kai version 1.1 divided the '''MCP Servers''' settings into two categories: Built-in and Custom. Now, the built-in MCP tools have a separate settings page that lists all tools with their names and descriptions. Additionally, a separate settings page for adding servers and customizing properties.

{{Menu|Tools > Options > Kai > Chat > [[Kai Options: Built-in|Built-in]]}}

Use the built-in settings to see a list of all the built-in MCP endpoints and choose which to apply.

{{Menu|Tools > Options > Kai > Chat > [[Kai Options: Custom|Custom]]}}

Use the Custom settings to add servers and create custom properties.

==Enhancements and Improvements==
'''The following are more relevant changes to Kai version 1.1''':
* The loading of sessions is now done via ACP instead of from the local files.
** There are no preloaded sessions anymore, so the chat now starts empty.
** Added a new '''Import sessions''' button to import the sessions for the current working directory of the active engine. If there are none, a message is shown.
** Closing the tab will delete the session only if the agent supports it.
* Authentication in {{Menu|Tools > Options > Kai}} is now done via ACP (Agent Client Protocol). Clicking one of the buttons imports the sessions for the current working directory as new tabs.
** All of the available authentication methods are now pulled directly from the agent.
** Gemini now has OAuth support via the browser and 3 different API key methods.
** The Copilot login will now open a terminal window that displays the device code to enter in the browser.
* Improved local support via the new '''Native LM Studio ACP''' AI agent we are shipping with Kai. Despite the name, this agent works for both LM Studio and Ollama.

==See Also==
* [[Kai Overview]]
* [[Kai Installation]]

Kai Options: Custom

Eric.revollo — Tue, 08 Sep 2026 14:20:55 GMT

Eric.revollo imported Kai Options: Custom by file upload (3 revisions)

New page

{{Parent|Kai Options: MCP Servers|MCP Servers}}

{{Menu|Tools > Options > Kai > MCP Servers > Custom}}

Use this screen to add Servers and create customized properties.

[[File:Kai custom settings.png|800px|border|=link]]

== MCP Servers Options ==

The following options are available in '''MCP Servers''':

{| border="1" frame="hsides" rules="rows"
! '''Item'''
! '''Description'''
|- valign="top"
|
'''Servers'''
|
Displays the list of configured MCP servers.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
Add
|
Adds a new MCP server configuration.
|- valign="top"
|
Remove
|
Removes the selected MCP server configuration.
|}
|- valign="top"
|
'''Properties'''
|
Contains settings for the selected MCP server.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Name'''
|
Field to specify the MCP server name.
|- valign="top"
|
'''Type'''
|
Specify the MCP server type. Choose between Local, Remote, or Remote (SSE).
|- valign="top"
|
'''Start command'''
|
Specifies the command used to start the MCP server.
|- valign="top"
|
'''URL'''
|
Specify the URL address for the remote MCP server.
|- valign="top"
|
'''Tools'''
|
Displays the tools available for the selected MCP server.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
Add
|
Adds a tool to the selected MCP server.
|- valign="top"
|
Remove
|
Removes the selected tool from the MCP server.
|}
|- valign="top"
|
'''Headers'''
|
Define Headers for a remote MCP Server.
{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Key'''
|
Specifies the environment variable name.
|- valign="top"
|
'''Value'''
|
Specifies the environment variable value.
|}
|- valign="top"
|
'''Environment variables'''
|
Defines environment variables used by the MCP server.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Key'''
|
Specifies the environment variable name.
|- valign="top"
|
'''Value'''
|
Specifies the environment variable value.
|}
|- valign="top"
|
'''Arguments'''
|
Displays the command-line arguments used by the MCP server.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
Add
|
Adds a command-line argument.
|- valign="top"
|
Remove
|
Removes the selected command-line argument.
|}
|}
|}

Kai can use configured MCP servers to access additional tools and services.

==See Also==
* [[Kai Options: Chat]]
* [[Kai Options: Completions]]
* [[Kai Options: ACP Registry]]
* [[Kai Features]]

Kai Options: Built-in

Eric.revollo — Tue, 08 Sep 2026 14:20:55 GMT

Eric.revollo imported Kai Options: Built-in by file upload (5 revisions)

New page

{{Parent|Kai Options: MCP Servers|MCP Servers}}

{{Menu|Tools > Options > Kai > MCP Servers > Built-in}}

MCP servers can provide external tools and integrations that extend Kai functionality. This configuration screen displays the entire list of built-in IDE MCP tools. Here, users can manually select or deselect which ones to apply.

[[File:Kai built-in.png|800px|border|=link]]

The following options are available:

{| border="1" frame="hsides" rules="rows"
! '''Item'''
! '''Description'''
|- valign="top"
|
'''Select All'''
|
Selects all available built-in options.
|-
|
'''Deselect All'''
|
Deselects all available built-in options.
|-}

Kai Options: MCP Servers

Eric.revollo — Tue, 08 Sep 2026 14:20:55 GMT

Eric.revollo imported Kai Options: MCP Servers by file upload (12 revisions)

New page

{{Parent|Kai Options}}

{{Menu|Tools > Options > Kai > MCP Servers}}

MCP servers can provide external tools and integrations that extend Kai functionality.

Configure additional MCP servers for Kai, either by using the built-in ones or by customizing new ones.

==Child Pages==
* [[Kai Options: Built-in]]
* [[Kai Options: Custom]]

==See Also==
* [[Kai Options: Chat]]
* [[Kai Options: Completions]]
* [[Kai Options: ACP Registry]]
* [[Kai Features]]

Kai Options: ACP Registry

Eric.revollo — Tue, 08 Sep 2026 14:20:55 GMT

Eric.revollo imported Kai Options: ACP Registry by file upload (6 revisions)

New page

{{Parent|Kai Options: Chat|Kai Chat}}

{{Menu|Tools > Options > Kai > Chat > ACP Registry}}

With ACP Registry support, users no longer need to install the Claude Agent ACP manually via NPM.

Kai now supports a much wider range of agents, which users can manually install from the ACP Registry configuration page. The ACP Registry currently includes 35+ agents: all the agents Kai previously supported remain available, and additional well-known agents have been added, including Cursor, Devin, Kilo Code, and Junie.

The ACP Registry configuration window displays a list of agents that users can manually '''Install''', '''Uninstall''', or '''Update'''.

[[File:Kai ACP Registry.png|link=|border|800px]]

{{Attention| Zed Industries and JetBrains maintain the registry, and Embarcadero has no control over or responsibility for the registry or the agents downloaded from it.}}

Other agents are only available to download via NPX (node.js) or UVX (PyPi Python). To install those agents, it is necessary to manually configure the computer to support the underlying technology.

==See Also==
* [[Kai Options: Chat]]
* [[Kai Version 1.1#ACP Registry Support|ACP Registry Support]]
* [https://agentclientprotocol.com/get-started/registry ACP Registry - Agent Client Protocol]
* [[Kai Options: Completions]]
* [[Kai Features]]

Kai Options: Chat

Eric.revollo — Tue, 08 Sep 2026 14:20:54 GMT

Eric.revollo imported Kai Options: Chat by file upload (10 revisions)

New page

{{Parent|Kai Options}}

{{Menu|Tools > Options > Kai > Chat}}

Use this page to configure AI chat providers for Kai.

Kai supports multiple AI chat providers, including cloud-based services and local models.

[[File:Kai chat v1.1.png|border|900px]]

==Child Pages==
* [[Kai Options: MCP Servers]]
** [[Kai Options: Built-in]]
** [[Kai Options: Custom]]

== Chat Page ==
The following options are available on this page. Keep in mind that the options change depending on the selected provider.

[[File:Kai chat crop.png|700px|border|link=]]

===GitHub Copilot Tab===
{{Menu|Tools > Options > Kai > Chat > GitHub Copilot}}

To enable GitHub Copilot as an agent provider for Kai Chat services, users must first log in to their accounts.

{| border="1" frame="hsides" rules="rows"
! '''Item'''
! '''Description'''
|- valign="top"
|
Status
|
Displays the current connection or authentication status for the selected provider. Green displays when the user is already signed in. The status displays Red if the user is not logged into their account.
|- valign="top"
|
Sign in
|
Opens the authentication process for the selected provider. A browser opens, requesting users to sign in.
|- valign="top"
|
Sign out
|
Signs out from the current provider.
|}

===OpenAI Codex Tab===
{{Menu|Tools > Options > Kai > Chat > OpenAI Codex}}

To enable OpenAI Codex as an agent provider for Kai Chat services, users must first log in to their accounts.

{| border="1" frame="hsides" rules="rows"
! '''Item'''
! '''Description'''
|- valign="top"
|
Status
|
Displays the current connection or authentication status for the selected provider. Green displays when the user is already signed in. The status displays Red if the user is not logged into their account.
|- valign="top"
|
Sign in
|
Opens the authentication process for the selected provider. A browser opens, requesting users to sign in.
|- valign="top"
|
Sign out
|
Signs out from the current provider.
|}

===Google Gemini Tab===
{{Menu|Tools > Options > Kai > Chat > Google Gemini}}

Kai uses a Bring Your Own Key (BYOK) model.
To activate Google Gemini for the chat, you first need an API key. Users can generate API keys using [https://aistudio.google.com/ Google AI Studio].

{| border="1" frame="hsides" rules="rows"
! '''Item'''
! '''Description'''
|- valign="top"
|
API key
|
Enter the API key.
|}

{{Note|When enabling Gemini for chat, sometimes it is necessary to restart {{Product}} or even maybe the PC for the API key to work correctly.}}

===Local Model Tab===
{{Menu|Tools > Options > Kai > Chat > Local Model}}

{| border="1" frame="hsides" rules="rows"
! '''Item'''
! '''Description'''
|- valign="top"
|
URL
|
Specifies the URL address used to connect to the selected local model provider.
|- valign="top"
|
Test Connection
|
Button to test the connection to the selected local model provider.
|- valign="top"
|
Name
|
Specifies the local model name.
|}

===Claude Code Tab===
For Claude Code users must login separetly.

==See Also==
* [[Kai Options: Completions]]
* [[Kai Options: MCP Servers]]
* [[Kai Features]]

Kai Options: Completions

Eric.revollo — Tue, 08 Sep 2026 14:20:54 GMT

Eric.revollo imported Kai Options: Completions by file upload (14 revisions)

New page

{{Parent|Kai Options}}

{{Menu|Tools > Options > Kai > Completions}}

Here, you can configure AI code-completion providers for Kai.

Kai supports multiple completion engines, including cloud-based providers and local models. You must sign in or configure the selected provider before using AI completions.

[[File:Completions local models.png|900px|border]]

== Completions Options ==

The following options are available in '''Completions''':

{| border="1" frame="hsides" rules="rows"
! '''Item'''
! '''Description'''
|- valign="top"
|
'''General'''
|
Contains general AI completion settings.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Enable auto completions'''
|
Enables or disables AI code completions in the Code Editor.
|- valign="top"
|
'''Show as'''
|
Selects how AI completions are displayed in the editor.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Ghost text'''
|
Displays inline completion suggestions directly in the editor.
|- valign="top"
|
'''CodeInsight symbol'''
|
Displays completion suggestions using a symbol indicator.
|}
|- valign="top"
|
'''Active model'''
|
Selects the AI provider used for code completions.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''GitHub Copilot'''
|
Uses GitHub Copilot for code completions.
|- valign="top"
|
'''Local model'''
|
Uses a locally installed LLM through supported local providers.
|}
|}
|- valign="top"
|
'''Model settings'''
|
Contains settings for the selected AI provider.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''GitHub Copilot'''
|
Displays settings for GitHub Copilot. Since this is an online model, you must first sign in.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Status'''
|
Displays the current connection or authentication status.
|- valign="top"
|
'''Message'''
|
Displays provider-related status or authentication messages.
|- valign="top"
|
'''Sign in'''
|
Opens the authentication process on a browser.
|- valign="top"
|
'''Sign out'''
|
Signs out from the selected provider.
|}
|- valign="top"
|
'''Local model'''
|
Displays settings for the selected local model provider.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Backend'''
|
Selects the backend used for the chosen local model provider.

{| border="1" frame="hsides" rules="rows"
|- valign="top"
|
'''Default backend'''
|
Uses the default local model backend.
|- valign="top"
|
'''Other (e.g. LM Studio)'''
|
Uses another compatible local model backend, such as LM Studio.
|}
|- valign="top"
|
'''URL'''
|
Specifies the URL address used to connect to the selected local model provider.
|- valign="top"
|
'''Test connection'''
|
Button to test the connection to the selected local model provider.
|- valign="top"
|
'''Name'''
|
Specifies the local model name.
|- valign="top"
|
'''Context size'''
|
Specifies the context size used by the local model.
|}
|}
|}

Kai uses a Bring Your Own Key (BYOK) model. Depending on the selected provider, you may need a valid account or API key.

==FIM Tokens Tab==
FIM (Fill-In-the-Middle) tokens are special control tokens that tell a language model to complete text between a given prefix and suffix, rather than simply continuing from the end of a prompt. They are particularly important for code completion in IDEs, where the model needs to insert code at the cursor while preserving the code that follows. Configuring these token parameters allows user integration with additional local AI engines.

Since different model families use different special tokens, Kai now lets you choose predefined options or manually customize the FIM tokens.

[[File:Kai completions fim.png|800px|border]]

{{Note|To determine if a local LLM supports FIM, check the model information on the backend model, such as Hugging Face or the vendor’s documentation. It often explicitly states “Supports Fill-in-the-Middle (FIM).”}}

Most popular local coding models with FIM support include:
* StarCoder
* CodeLlama
* DeepSeek-Coder
* Qwen2.5-Coder
* Codestral
* Replit Code models

==See Also==
* [[Kai Options: Chat]]
* [[Kai Options: MCP Servers]]
* [[Kai Features]]

Kai Options

Eric.revollo — Tue, 08 Sep 2026 14:20:54 GMT

Eric.revollo imported Kai Options by file upload (9 revisions)

New page

{{Parent|Tools Options Dialog Box}}

{{Menu|Tools > Options > Kai}}

==Active License==
If an active Kai license is found on the system, the following sub-pages define the Kai configuration options:
* [[Kai Options: Completions]]
* [[Kai Options: Chat]]
** [[Kai Options: ACP Registry]]
** [[Kai Options: MCP Servers]]
*** [[Kai Options: Built-in]]
*** [[Kai Options: Custom]]

[[File:Kai SettingsIDE.png|border|700px]]

== Unlincesed ==
If no Kai license is found, this page offers information on how to obtain and register a [[Kai Licensing|Kai license]].

== See Also ==
* [[Kai Overview]]

Using Local AI Models with Kai

Eric.revollo — Tue, 08 Sep 2026 14:20:53 GMT

Eric.revollo imported Using Local AI Models with Kai by file upload (24 revisions)

New page

{{Parent|Kai}}

Kai can connect to locally hosted Large Language Models (LLMs), allowing you to use AI-assisted development while keeping prompts and source code within an environment you control.

Using local models can help you with the following:
* Keeping prompts and source code within your local network.
* Avoiding model subscriptions and token-usage.
* Selecting a model that fits your hardware and development needs.
* Running the model on the same computer as {{Product}} or another computer on your network.
* Using separate models for [[Kai Options: Chat|Kai Chat]] and code completions.

The performance and capabilities of a local model depend on the model, the local model runtime, and the available hardware.
Kai version 1.0.2 expands local model support for [[Kai Features#Ghost Text|Ghost Text]], [[Kai Features#CodeInsight Completions|Code Completions]], and [[Kai Features#Suggestions|Code Suggestions]] by adding additional local backends and configurable Fill-In-the-Middle support.

==Topics==
* [[#Local Model Hardware Requirements|Local Model Hardware Requirements]]
* [[#Choosing a Local Model Runtime|Choosing a Local Model Runtime]]
* [[#Choosing a Local Model|Choosing a Local Model]]
* [[#Setting Up LM Studio|Setting Up LM Studio]]
* [[#Connecting Kai Chat to a Local Model|Connecting Kai Chat to a Local Model]]
* [[#Connecting to a Model on Another Computer|Connecting to a Model on Another Computer]]
* [[#Using a Local Model in Kai Chat|Using a Local Model in Kai Chat]]
* [[#Configuring Local Code Completions|Configuring Local Code Completions]]
* [[#Improving Local Model Results|Improving Local Model Results]]
* [[#Troubleshooting|Troubleshooting]]
* [[#Security Considerations|Security Considerations]]

== Local Model Hardware Requirements ==
Kai does not impose a fixed hardware requirement for local models. The required resources depend on the following:
* The model size.
* The selected quantization.
* The configured context-window size.
* The local model runtime.
* Whether the model uses CPU or GPU acceleration.
* Whether {{Product}} and the model runtime run on the same computer.

The following recommendations apply to the computer hosting the local model, not only to the computer running {{Product}}.

==Choosing a Local Model Runtime==
A local model runtime downloads, loads, and runs the model. It also provides the API endpoint that Kai uses to connect.

Common options include LM Studio, Ollama, llama.cpp, and oMLX.

===LM Studio===
[https://lmstudio.ai LM Studio] provides a graphical interface for:
* Finding and downloading models.
* Loading and unloading models.
* Starting a local API server.
* Viewing the server URL.
* Allowing access from other computers on the local network.

LM Studio is a practical choice when you prefer a graphical interface.

===Ollama===
[https://ollama.com Ollama] provides command-line tools and a local API for downloading and running supported models.

Follow these steps to prepare Ollama for Kai:
# Install Ollama.
# Download a supported coding model.
# Start the Ollama service.
# Identify the API URL and port.
# Enter the URL in Kai.
# Test the connection.
# Select the model.

LM Studio and Ollama use different ports. Always use the URL displayed or documented by the runtime.

===llama.cpp===
The [https://llama-cpp.com llama.cpp] model provides greater control over model loading and resource use, but it usually requires more manual configuration.

It is best suited to advanced users who need a customized local model environment.

===oMLX===
[https://omlx.ai oMLX] provides a native macOS interface for:

* Finding and downloading models optimized for Apple Silicon.
* Loading and managing models in MLX format.
* Starting an OpenAI-compatible local API server.
* Viewing the server URL and port.
* Using the Mac’s unified memory and MLX hardware acceleration.
* Allowing RAD Studio in a Windows virtual machine or another computer on the local network to connect to the model.

oMLX is a practical choice when you use an Apple Silicon Mac and want models specifically optimized for Apple’s MLX framework.

==Choosing a Local Model==
Not every model is suitable for software development or for use with Kai.

Consider the following when selecting a model:
* Whether it is designed for coding.
* Whether it supports tool use for Kai Chat.
* Whether it supports Fill-In-the-Middle for code completions.
* Whether it fits within the available memory.
* Whether it is compatible with your selected runtime.
* Whether its response speed is appropriate for your workflow.

A model that works well for Kai Chat might not be suitable for inline code completion.

===Choosing a Model for Kai Chat===
For Kai Chat and agentic development tasks, select a coding model that:
* Supports tool use.
* Has a context window suitable for the project.
* Fits comfortably within the available memory.
* Is compatible with the selected runtime.

Common coding model families available through local runtimes include:
* Qwen Coder.
* Gemma coding-capable models.
* DeepSeek coding models.
* Smaller coding models for systems with limited memory.

Test the model with tasks representative of your own projects, such as:
* Creating a Delphi class.
* Explaining an existing unit.
* Refactoring a method.
* Generating a DUnitX test.
* Correcting a compiler error.

===Choosing a Model for Code Completions===
For Ghost Text, Code Completions, and Code Suggestions, select a coding model that explicitly supports Fill-In-the-Middle, or FIM.

FIM allows the model to generate code at the cursor using both the code before and after the insertion point

Common [[Kai Options: Completions|FIM-capable]] model families include:
* StarCoder.
* Code Llama.
* DeepSeek-Coder.
* Qwen2.5-Coder.
* Codestral.
* Replit Code models.

Verify FIM support for the exact model version you download.

===Using LLM Fit===
[https://github.com/AlexsJones/llmfit LLM Fit] evaluates your computer and identifies models that are likely to run within the available resources.

Follow these steps to use LLM Fit:
# Install LLM Fit from its project repository.
# Open a terminal or command prompt.
# Run:
# Review the detected hardware and available memory.
# Filter the results by the coding use case.
# Review the estimated model memory use and speed.
# Select a model that fits comfortably within the available resources.
# Download the corresponding model through LM Studio, Ollama, or another supported source.

Avoid models identified as marginal or borderline unless you are prepared for reduced performance or memory errors.

==Setting Up LM Studio==
Follow these steps to prepare LM Studio for use with Kai.
# Download LM Studio from [https://lmstudio.ai/ https://lmstudio.ai/]
# Install and open LM Studio.
# Search for a coding model.
# Select a model size and format appropriate for your hardware.
# Download the model.
# Load the downloaded model.
# Open the local server area.
# Start the local API server.
# Record the server URL displayed by LM Studio.

A typical local URL is:
<nowiki>http://localhost:1234/v1/</nowiki>

The exact URL may vary.

===Enable Network Access===
Enable network access when {{Product}} and the model runtime are running on different computers or when RAD Studio is running in a virtual machine.
To enable access:
# Open the LM Studio server settings.
# Enable access from the local network.
# Record the host computer’s IP address or network name.
# Confirm that the operating-system firewall allows access to the server port.

{{Note|Use this option only on a trusted network.}}

==Connecting Kai Chat to a Local Model==
Follow these steps to connect Kai Chat to LM Studio, Ollama, or another compatible local server:
# Start the local model runtime and load or activate the coding model.
# Start the runtime’s API server.
# Open {{Product}} and select {{Menu| Tools > Options}}.
# In the Options window, select {{Menu| Kai > Chat}} to open the Chat settings.
# Select the '''Local Model''' tab.
# Enter the server URL.
# Test the connection with the'''Test Connection''' button to confirm that the connection succeeds.
# Select the model.
# Save the settings.

When the connection succeeds, Kai displays the models exposed by the local runtime.

{{Note|A model configured for Kai Chat is not automatically used for code completions.}}

==Connecting to a Model on Another Computer==
It is possible to run the model on another computer on the same trusted network. You can also run a model on a cloud-hosted machine; in that case, {{Product}} recommends using an API key for additional security.

This configuration is useful in the following situations:
When the {{Product}} computer has limited system or GPU memory.
Another computer has a more capable GPU.
{{Product}} runs in a virtual machine.
To prevent the model runtime from competing with {{Product}} for resources.

To connect to another computer, follow these steps:
# Start the model runtime on the host computer and load the required model.
# Start the API server.
# Enable network access in the runtime.
# Determine the host computer’s IP address or network name.
# Confirm that the firewall permits access to the API port.
# Open {{Product}} and navigate to the appropriate Kai settings.
# Enter the network URL and test the connection.
# Select the model and save the settings.

{{Note|Do not use {{code|localhost}} when the model runs on another computer. {{code|localhost}} refers to the computer running {{Product}}.}}

When {{Product}} runs in a virtual machine, use a host address that is reachable from the virtual machine. It is strongly recommended to enable an API key if your chosen inference engine supports it. For example, LM Studio has an entry for this. Then, in Kai, enter the API key. This helps prevent unauthorized use of your LLM.

==Using a Local Model in Kai Chat==
Once the model configurations are completed, follow these steps to use the local model in the Kai Chat:
# Open a project in RAD Studio.
# Open the Kai Chat window.
# Create a new local session.
# Select one of the available models from the model combobox (in the bottom panel).
# Enter a clear development request.
# Approve or reject requested operations, if requested.
# Review the files created or modified by Kai.
# Compile and test the changes.

Additionally, local models can also help you:
* Explain existing code.
* Create classes and units.
* Refactor methods.
* Generate tests.
* Diagnose compiler errors.
* Add validation and error handling.
* Generate documentation.

{{Note|Ensure to always review and test generated code before using it in a production project.}}

===Manage Permissions===
Kai may request permissions before any of the following actions:
* Reading project files.
* Searching the project.
* Creating or modifying files.
* Accessing a directory.
* Using an MCP tool.

Use restricted permissions when evaluating a new model. Keep the project under source control and review all generated changes.

==Configuring Local Code Completions==
In version 1.0.2, Kai expands local support for [[Kai Features#Ghost Text|Ghost Text]], [[Kai Features#CodeInsight Completions|Code Completions]], and [[Kai Features#Suggestions|Code Suggestions]].

For local completions, Kai allows users to perform the following configurations:
* The local server URL.
* The backend engine.
* The model.
* The FIM token format.

===Before Configuring Completions===
Before configuring a local completion model, ensure to complete this set of steps:
* Install Kai version 1.0.2 or later.
* Install a compatible local model runtime.
* Download a FIM-capable coding model.
* Load the model.
* Start the local API server.
* Record the server URL.

===Configure the Completion Model===
Follow these simple steps to configure a local completion model.
* Open {{Product}} and navihgate to {{Menu|Tools > Options > Kai}}.
* In the Kai window, select the '''Completions''' settings.
* Select the '''Local Models''' tab.
* Enter the local server URL.
* Select the backend engine.
* Select the model.
* Select the FIM configuration that matches the model.
* Save the settings.
* Open a Delphi or C++ source file.
* Begin coding and verify that a completion appears.

Different backends can expose and process models differently. Select the backend that corresponds to the runtime used to serve the model.

===Configure FIM Tokens===
Kai provides predefined FIM configurations for supported model families.

Use a predefined option when it matches the selected model.

When no matching preset is available:
* Review the documentation for the exact model version.
* Identify the model’s prefix, suffix, and middle tokens.
* Open {{Menu|Tools > Options > Kai > Completions}}.
* Select '''Local Models'''.
* Select the server, backend, and model.
* Select the manual FIM option.
* Enter the required tokens.
* Save and test the configuration.

Incorrect FIM tokens can cause the model to repeat existing code, continue from the wrong location, or return unusable suggestions.

===Test Code Completions===
Follow these steps to test the configuration:
# Open a Delphi or C++ source file.
# Place the cursor inside a method or function.
# Begin typing a statement or code block.
# Pause briefly for a suggestion.
# Review the Ghost Text or completion.
# Accept or dismiss the suggestion.
# Repeat the test in several code contexts.

For inline completion, response speed is especially important. A smaller FIM-capable model may provide a better editing experience than a larger chat-oriented model.

===Use Different Models for Chat and Completions===
You can configure a larger, tool-capable model for Kai Chat, or a smaller, faster FIM-capable model for code completions.

When system memory is limited, avoid loading two large models simultaneously.

==Improving Local Model Results==
This section explains how to improve local model results.

'''Provide Specific Instructions'''

Ensure to provide specific instructions that include the following information:
* The language and framework.
* The required file or unit name.
* The classes and methods to create.
* Validation and error-handling requirements.
* Files that must not be modified.
* Required tests.
* The targeted {{Product}} version.

'''Break large tasks into Stages'''

Try breaking large tasks into smaller stages for a complex change:
# Ask Kai to inspect and summarize the existing code.
# Request a proposed plan.
# Review the plan.
# Apply one stage at a time.
# Compile and test after each stage.

'''Start a New Session when necessary'''

Start a new session when:
* The conversation becomes very long.
* The model stops following earlier instructions.
* You move to an unrelated task.
* You want to provide a clean set of requirements.

==Troubleshooting==
This section explains the most common issues and how to troubleshoot them.

===Test Connection Fails===
If the test connection fails, verify the following to troubleshoot the issue.
* The model runtime is running.
* The API server has started.
* The model is loaded.
* The URL and port are connected.
* The firewall allows access.
* Enable Network access for remote connections.
* You are not using {{code|localhost}} for a remote model.

===Kai does not display any Models===
If Kai is not displaying any models, verify that:
* A model has been downloaded.
* The model is available to the API server.
* The correct backend is selected.
* The connection test succeeds.

Restart the API server if the model list does not refresh.

===The Model is Slow or Runs Out of Memory===
If the model is runnig slow of runs out of memory, try the following troubleshooting:
# Selecting a smaller model.
# Using a more efficient quantization.
# Reducing the context-window size.
# Closing memory-intensive applications.
# Confirming that GPU acceleration is active.
# Running the model on another computer.
# Using LLM Fit to identify a better model.

===Chat Works, but Code Completions Does Not===
If the chat is working but the code completion feature is not, verify the following to troubleshoot the issue.
* The completion model supports FIM.
* The correct backend is selected.
* The correct FIM preset is selected.
* Manually entered FIM tokens match the exact model version.
* The '''Local Models''' option is selected in the Kai Completions settings.

{{Note|Chat support does not guarantee FIM support.}}

===Suggestions Repeat Existing Code or Appear Misplaced===
The Suggestions feature may be repeating existing code or appearing in incorrect places.

Check if the FIM is configured correctly and verify the selected preset or the manually entered prefix, suffix, and middle tokens.

===The Model Produces Incorrect Code===
# If the model is returning incorrect, try the following:
# Specify the exact RAD Studio version.
# Identify Delphi or C++Builder.
# Identify VCL or FireMonkey.
# Provide the relevant existing declarations.
# Compile the code.
# Return the exact compiler error to Kai.
# Ask Kai to correct only the reported issue.

==Security Considerations==
A local model can keep prompts and code within the infrastructure under your control, but the complete configuration may include other tools and services. For better security measurements, review the following:
* Whether the runtime sends optional telemetry.
* Whether MCP servers connect to external systems.
* Whether the API server is accessible outside the intended network.
* Whether downloaded models come from trusted sources.
* Whether project files contain credentials or regulated data.

Do not expose an unauthenticated local-model server directly to the public internet. Instead, use the following:
* A trusted private network.
* Appropriate firewall rules.
* Authentication, when supported.
* Source control to review and reverse generated changes.
* Least-privilege permissions for project and file access.

== See Also ==
* [[Kai Installation]]
* [[Kai Options]]
* [[Kai Options: Chat]]
* [[Kai Options: Completions]]
* [[Kai Options: MCP Servers]]
* [[GetIt Package Manager]]

Kai Licensing

Eric.revollo — Tue, 08 Sep 2026 14:20:53 GMT

Eric.revollo imported Kai Licensing by file upload (14 revisions)

New page

{{Parent|Kai}}

This section focuses primarily on the process to obtain a license for Kai, Embarcadero’s new agentic AI for {{Product}}. Ensure to first download Kai from the [[GetIt Package Manager]], or read more in the [[Kai Installation]] page.

== Obtain a Kai Trial License ==
A specific license is needed to use Kai features. This is a separate license from {{Product}}, Delphi, or C++. Users with an existing RAD Studio, Delphi, or C++Builder 12.0 or higher license can request a 30-day free trial key. To do so, visit the [https://reg.embarcadero.com/srs6/promotion.jsp?promoId=590 BOGO redemption] page and enter your existing {{Product}} key to obtain a Kai trial key.

{{Note|Users who do not have {{Product}} 13.1 installed, requesting a {{Product}} 13.1 trial includes a license for Kai out of the box.}}

[[File:Trial license.png|border]]

== Buy a Kai Subscription ==
Users can buy a Kai subscription along with a new {{Product}}, Delphi, or C++Builder license, either by selecting the option in the [https://www.embarcadero.com/app-development-tools-store online shop] or by contacting support when purchasing a new license.

For users with an active subscription, contact support and request a new Kai license.

Users with an existing {{Product}} license but no active subscription must contact support to activate the subscription and acquire the Kai subscription.

Once a user acquires a Kai subscription, the user will receive an email with the license key.

== Register the Kai License ==
Once a user acquires a trial or a subscription license key for Kai, the next step is to register it. To register the new license, navigate to the bin subfolder under the {{Product}} installation folder ({{code|C:\Program Files (x86)\Embarcadero\Studio\37.0\bin}}) and start the {{code|LicenseManager.exe}} application.

[[File:Kai license.jpg|border|800px]]

=== Active License ===
Check if the license is registered and active by checking the {{Product}} splash screen. The Kai widget indicates if the license is active, not found, or expired.

[[File:Kai license status.png|border|700px]]

For active licenses, users can set and customize the Kai configuration in {{Product}} by navigating to {{Menu|Tools > Options > [[Kai Options|Kai]]}}.

[[File:Kai SettingsIDE.png|border|650px]]

== See Also ==
* [[Kai Overview]]
* [[Kai Features]]
* [[Kai Installation]]

Kai Features

Eric.revollo — Tue, 08 Sep 2026 14:20:53 GMT

Eric.revollo imported Kai Features by file upload (21 revisions)

New page

{{Parent|Kai}}

This section explains in more detail the different Kai UI Features.

==Ghost Text==
The Ghost text feature introduces automatic, context-aware suggestions that appear directly in the editor as faint, translucent inline text, providing a fast, unobtrusive way to complete code.

Ghost text suggestions are triggered automatically after a brief pause in typing, helping users maintain focus and coding flow. Accept a suggestion instantly by pressing the '''Tab''' key, or dismiss it with the '''Esc''' key.

[[File:Ghost-Text-Light.png|600px|border]]

==CodeInsight Completions==
Kai enhances [[Source Options#Code Insight tab|CodeInsight]] by integrating intelligent typing suggestions directly into the {{Product}} completion window. These suggestions appear as the top item in the completion list, making them immediately accessible.

Users can invoke the completion list with {{code|Ctrl+Space}} and accept suggestions with {{code|Enter}}.

[[File:CodeInsight-Completions-Light.png|600px|border]]

==Suggestions==
Kai also introduces '''Suggestions''', a dedicated panel that lets users generate and compare multiple code suggestions at once. Accessible via {{Menu|View > Kai Suggestions}}, this view facilitates the evaluation of alternative implementations, enabling users to choose the most suitable option.

Users can manually refresh the panel to generate new suggestions, and each option includes convenient actions to either copy the code or insert it directly into your editor.

[[File:Panel-Completions-Light.png|800px|border]]

{{Note|Kai Suggestions are powered by the active completion engine in the configuration (like the other completion features).}}

==Agent Chat==
Users can now use the Kai Agent directly in the IDE as a fully integrated assistant for coding tasks, technical questions, and project analysis. Accessible via {{Menu|View > Kai Chat}}, it enables you to interact with AI to compile, debug, and implement features without leaving your development environment.

The use of an Agentic Chat provides the following advantages:
* MCP-powered Agentic Workflows that handle files, resolve build errors, and manage version control.
* Perform File System Operations such as Read, Write, and manage project files.
* Automate a Compile-Fix Workflow by allowing Kai to compile an application, read errors, attempt fixes, and recompile automatically.
* Version Control Integration allows users to create and manage pull requests directly from the IDE.

=== Choose an Engine===
When users start a chat, they can select which engine to use for the session. These engines must be configured in the [[Kai Options: Chat|Kai Chat Settings]] beforehand.

Open a new chat by clicking the new message icon [[Image:New chat icon.png|25px]], and select one of the available LLMs using the drop-down menu to start a chat session.

[[File:Agent-Chat-Providers-Light.png|border|500px]]

Depending on the chosen LLM engine, users can select the agent model. Some of the mode types include additional parameters such as, '''Agent''' mode, '''Read Only''' mode, or '''Full Access''' mode with the ability to make code changes. The options depend on the selected agent.

{{Note|There are also additional engine-specific configurations in additional combo boxes. Each combo box shows its role in a description that appears when the mouse hovers over it.}}

Now, you can start typing instructions for Kai to help create code for your application.

[[File:Agent-Chat-Light.png|border|500px]]

The section below shows examples of agent models for each agent and their options.

'''Claude Code Agent Sample''':

[[File:Agent-Chat-Claude-Light.png|border|500px|=link]]

'''Codex Agent Sample''':

[[File:Agent-Chat-Codex-Dark.png|border|500px|=link]]

'''Copilot Agent Sample''':

[[File:Agent-Chat-Copilot-Light.png|500px|border|=link]]

'''Gemini Agent Sample''':

[[File:Agent-Chat-Gemini-Dark.png|500px|border|=link]]

===Tabs for Chats ===
Users can have multiple tabs; each tab hosts a chat. These chats persist across IDE sessions. It is possible to chat with different engines at the same time using different tabs.

Close one tab with the {{code|X}} symbol, or close all tabs using the fire icon [[File:Close-chat-icon.png|25px]] from the toolbar.

===Additional Features===
====Chat File Reference====
Refer to files in the chat by using the {{code|@}} key and typing the filename.

[[File:Kai file ref.png|400px|border]]

====Chat Slash Commands====
Users can start typing the {{code|/}} symbol to get useful command suggestions; each agent has its own specific slash commands.

[[File:Agent-Chat-Slash-Commands-Dark.png|500px]]

Users can locally define agent skills to add as slash suggestions.

==See Also==
* [[Kai Overview]]
* [[Kai Options]]
* [[Kai Licensing]]

Kai Release Notes

Eric.revollo — Tue, 08 Sep 2026 14:20:52 GMT

Eric.revollo imported Kai Release Notes by file upload (6 revisions)

New page

{{Parent|Kai}}

These Release Notes contain important Kai information based on each {{Product}} version.

== Kai version 1.1 ==
The Kai version 1.1 release changes how Kai interacts with the various AI agents and local CLI engine.

The following are the core new features for Kai 1.1:
* '''ACP Registry Integration''': The integrations with the different CLI AI engines don't ship with Kai anymore, but they are downloaded after the installation from a public registry. This registry provides access to dozens of AI integrations, offers a direct update mechanism when new releases become available, and will add new options as they are added to the registry, without requiring a Kai update.
* '''MCP Servers Update''': This release divided the MCP Servers configuration settings into two categories: Built-in and Custom options, where users can select their preferred IDE MCP tools.
* '''UX Improvements''': Kai 1.1 has a large number of UX and UI improvements across all screens to bring it more in line with RAD Studio UI and standard behavior.
* '''Quality Enhancements''': The new Kai version improves quality and addresses nuances and issues, specifically focused on the chat window.

{{Note|For more information on these features, see the [[Kai Version 1.1]] documentation.}}

== Kai version 1.0.2 ==
The Kai 1.0.1 update focuses on improving compatibility, installation reliability, internationalization, and overall product quality.

This version update primarily addresses international language support and other enhancements, including support for completions using additional AI engines.

{{Note| {{Product}} strongly recommends that all Kai users install this update, regardless of whether they use a paid license or a trial version.}}

=== Expanding Completions Support ===
This release expanded Kai completions (Ghost test, Code Completions, and Code Suggestions) to integrate with additional local AI engines.

With this update, you can make different engines work by configuring a new Kai parameter: FMI tokens. Navigate to {{Menu|Tools > Options > Kai > [[Kai Options: Completions|Completions]]}}, select a local model tab, and select the URL as before, but now users can also pick the backend engine to use. Because different model families use different special tokens, Kai now lets you pick predefined options or manually customize the FIM tokens.

=== Quality and Reliability Improvements ===
The following are the main improvements:
* Better international language support. Improved compatibility with non-English Windows environments, including systems using ANSI code pages, extended European characters, and Japanese and Korean input methods (IME), making Kai work more naturally for developers around the world.
* Expanded support for additional AI deployment scenarios, including improvements for LiteLLM environments and GitHub Enterprise Cloud authentication when using GitHub Copilot.
* Improved usability by adding small enhancements to everyday workflows, such as the ability to copy content directly from the Kai chat window and improved handling of Windows user profiles containing spaces.
* This release includes internal security improvements that strengthen Kai’s MCP communication infrastructure.

== Kai version 1.0.1 ==
The Kai 1.0.1 update focuses on quality, stability, and usability improvements. The following are the new features and quality improvements for this version.

{{Note| {{Product}} strongly recommends that all Kai users install this update, regardless of whether they use a paid license or a trial version.}}

=== Authenticated Connections for Local AI Engines ===
The 1.0.1 update adds support for authenticated local AI servers, allowing Kai to authenticate when connecting to them, such as Ollama and LM Studio. This improves security by enabling secure connections to AI models hosted on remote or cloud-based infrastructure.

=== Quality and Stability Improvements ===
The following are the main improvements:
* Improved stability and reliability.
* Improved authentication and connectivity for both local and cloud AI models, including Ollama, LM Studio, and Claude.
* Improved inline completions, ghost text, suggestion rendering, clipboard operations, and compatibility with both modern and Classic Code Insight.
* Visual and user interface issues have been addressed, including layout, theming, font sizing, icon rendering, resizing behavior, navigation controls, and clearer terminology.

==Kai Known Issues in {{Product}} Versions==

==={{Product}} Florence 13.1===
'''Stability Issues with Gemini version''':

The {{code| 0.42.0-preview.0}} release of Google Gemini CLI is included as part of the Kai 1.0 release. Reports indicate stability issues in some environments with this version of the Google Gemini CLI.

Customers experiencing stability-related behavior can manually downgrade to version {{code|0.35.3}} to mitigate the issue in affected environments. However, downgrading may reintroduce security vulnerabilities or dependency issues that were addressed in later releases, such as the {{code|0.42.0-preview.0}} release.

==See Also==
* [[Kai Overview]]
* [[Kai Features]]

Kai Installation

Eric.revollo — Tue, 08 Sep 2026 14:20:52 GMT

Eric.revollo imported Kai Installation by file upload (10 revisions)

New page

{{Parent|Kai}}

This section explains how to download and install Kai for {{Product}}.

== Download Kai ==
Download Kai using the [[GetIt Package Manager]] in any edition of {{Product}} 13 Florence, Delphi 13 Florence, and C++Builder 13 Florence. This includes both versions 13.0 and 13.1, as well as the Trial SKU.

Users can also install Kai in {{Product}} 12.x, Delphi 12.x, and C++Builder 12.x (32-bit IDE only), except for the Community Editions.

Using the GetIt package manager, search for '''Kai''' and click '''Install'''.

[[File:Kai getit package.png|650px|border|=link]]

Another option is to download Kai from the [http://my.embarcadero.com/ Embarcadero customer portal] with an active EDN account.

{{Note|The latest Kai version 1.0.2, is available in GetIt. You can find it in the [[GetIt Package Manager]] dialog or from the Kai Welcome Page. The installer automatically removes the previous version and installs Kai 1.0.2. A restart of RAD Studio is required to complete the update.}}

==Kai Installation==
Once downloaded from the GetIt package manager, the installation window displays. Follow the steps to complete the installation.
# Ensure to read the License Agreement information, select the '''I accept the agreement''' option, and click '''Next'''.
# Ensure to read the Information screen and click '''Next'''.
# Select the destination location folder.
# On the components screen, you can select components manually or choose one of the drop-down options: '''Full installation''', '''Compact installation''', '''Custom installation''', '''Local models only''', '''Chat only''', or '''Completions only'''.[[File:Kai install selectComponents.png|=link|border]]
# Select any additional task you wish to perform.
# On the Ready to Install screen, review all selected settings, then click '''Install''' to begin installation.

Once the installation is complete, the {{Product}} splash screen displays a specific entry for the Kai package, along with its [[Kai Licensing|licensing status]].

[[File:Kai license status.png|border|600px]]

Additionally, users can now customize and configure Kai Options, such as [[Kai Options: Completions|Completions]], the [[Kai Options: Chat|Chat]], and additional [[Kai Options: MCP Servers|MCP servers]], by navigating to {{Menu|Tools > Options > [[Kai Options|Kai]]}}.

===Installed CLIs===
The following CLI local LLMs interfaces are preinstalled with Kai. These tools are installed in the main Kai folder by default.

* Copilot Language Server
* Copilot CLI
* GitHub CLI
* Codex CLI
* Gemini CLI
* LLM-LS
* Codex ACP
* Claude Agent ACP

=== Claude Integration ===
If you wish to integrate Claude Code as an LLM interface, you first need to have npm installed to install the Claude Agent ACP. The next step would be to install Claude Code. Note that the Claude Agent is not the same as the Claude desktop app. You need to install the Claude Agent even if you have previously installed Claude Desktop.

{{Note| Claude Code cannot be legally distributed by a third-party installer.}}

Follow the steps below to integrate Claude Code:

1. Close {{Product}} if it is open.

2. Download the [https://nodejs.org/ Node.js LTS] installer.

3. Install Node.js; this also installs npm.

4. Open a terminal to use npm, and run:
Npm install -g @agentclientprotocol/claude-agent-acp

5. Install Claude Code from PowerShell and add it to the path as needed, for example:
Irm http://claude.ai/install.ps1 | iex

6. Now, from the terminal window, run the following Claude command:
claude

7. A prompt pops up asking users to log in to Claude. Follow the instructions and, in the browser window, log in to your Claude account. Once logged in, return to the terminal window.

{{Note| At this point, the installation is complete, but it is necessary to ensure that Claude tools are visible in {{Product}}. This can be done by checking the path environment variable.}}

8. Click the Windows key and search for "'''envir'''". Select the '''Edit the system environment variables''' option.

[[File:Env variables search.png|=link|border]]

9. When the System Properties window opens, ensure you are in the '''Advanced''' tab and click the '''Environment Variables''' option.

[[File:Env variables sysprop.png|=link|border]]

10. In the Environment Variables window, select the '''Path''' variable and click the '''Edit button'''.

[[File:Env variables path.png|=link|border]]

11. In the dialog box that appears, click on the '''New''' button.

[[File:Env variables.png|=link|border]]

12. Next, edit the new blank entry by typing:
%USERPROFILE%\.local\bin

13. Click '''Ok''' and close the Environment variable window.

14. Now, when you go into {{Product}}, you will be able to select Claude as one of the options in the Kai Chat window.

{{Note| When navigating to the {{Menu|Tools > Options > [[Kai Options|Kai]]}} settings, Claude Code cannot detect whether you are logged in; therefore, the Kai settings configuration window displays a message requesting that users log in. This is static and does not reflect the real status.}}

==Migration Guides==
This section details how to migrate versions in Kai.

===Migrating from Kai 1.0.x to Kai 1.1===
The introduction of the ACP registry and the agent upgrade can cause incompatibilities with previous versions of Kai. The following differences should be considered:
* For OpenAI Codex integration, if you see a compatibility error with SQLite, it’s because the database is incompatible. Delete the {{code|.codex}} folder under the user account and start anew.
* Some of the CLI agents use environment variables for configuration. These new environment variables are not generally visible to the {{Product}} IDE unless it is restarted. While this step is not required for every installation, it can be helpful in several specific scenarios (which are not always reproducible). If the new agents are not visible after configuration and login, consider restarting the {{Product}} IDE.
** An example is the Claude integration: after installing Claude support and logging in, it may be necessary to restart the IDE for the integration to work (the visible issue is that the chat is not responding).
** For OpenAI integration: if you see the error “{{code|CODE_API_KEY or OPENAI_API_KEY is not set}}”, this is an issue in the CLI (independent of Kai). It is recommended to set that environment variable manually.
** For GitHub: after installing GitHub’s Copilot AI tooling, you might see an error that no authentication exists until you restart the {{Product}} IDE.

==See Also==
* [[Kai Overview]]
* [[Kai Options]]
* [[Kai Licensing]]

Kai Overview

Eric.revollo — Tue, 08 Sep 2026 14:20:52 GMT

Eric.revollo imported Kai Overview by file upload (2 revisions)

New page

{{Parent|Kai}}

Kai is Embarcadero’s agentic AI platform for Delphi, C++Builder, and RAD Studio, bringing automation into your development workflow.

== Overview ==
Kai transforms RAD Studio into an agentic development environment, pairing best-in-class code suggestions with an AI agent that can generate, build, fix, and evolve your applications. More than a coding assistant, Kai acts as a collaborator: generating code and UIs, compiling projects, resolving errors, and executing tasks inside and outside the IDE using MCP-powered workflows.

From accelerating new applications to modernizing and refactoring existing codebases, Kai serves as an agentic companion, assisting with complex work and adapting to your process. Built on a flexible BYOK model, Kai supports leading LLMs, including GitHub Copilot, OpenAI, Claude, Gemini, and local models, giving developers control over performance, cost, and data privacy.

==Why Kai==
Kai allows users to stay in flow, to work entirely within {{Product}}. Easily turn prompts into working code, builds, and fixes without external disruptions.

Kai offers AI-powered inline autocomplete and rich code suggestions to improve coding and boost productivity. This easy-to-use tool lets users generate UIs, working prototypes, or even apps by describing an idea, then have Kai create a ready-to-use platform or a functioning application.

Users with existing applications can use Kai to modernize the code and UI, and help refactor existing codebases. Using the best model for specific tasks and leveraging new language and framework features.

When it comes to security measures, Kai does not use any proxy layer via Embarcadero services, nor does it have any mandatory cloud dependencies; your code stays under your control through local storage and direct model communication. In addition, Kai can be run offline with local LLMs for maximum IP protection and compliance.

== Kai Architecture ==
Kai features an integrated AI agent that interacts directly with the development environment. Kai connects to leading AI models through a flexible BYOK (Bring Your Own Key) approach, supporting providers such as GitHub Copilot, OpenAI, Anthropic, Google, and various local LLMs.

It interfaces with local CLI engines via an MCP (Model Context Protocol) layer, enabling not only interaction with the IDE itself but also integration with external MCP servers.
The Kai tool is installed as an IDE package that provides IDE integration and a conduit for LLM integration. The following list of CLI tools is used:
* github/copilot-language-server
* github/copilot
* openai/codex
* google-gemini/gemini
* Codex and Cloud Agent ACP interfaces

Each of the tools mentioned above offers integration with multiple models. The solution can also integrate local LLMs. The image below displays how Kai integrates with {{Product}} and the local LLMs.

[[File:Kai architecture.png|800px|border]]

==MCP Support==
MCP support is an integral part of the solution. When the tool makes a request, it also provides endpoints in the IDE so that external AI engines can interact with the IDE in multiple ways, with instructions detailed in the AGENTS.md file, passed along with each request.

=== MCP Agentic Workflows ===
Kai includes an MCP server interface that lets users integrate other MCP tools. The MCP server understands the file system, allowing users to read, write, and manage project files without extra configurations.

Kai provides an MCP interface for {{Product}} that automates compile-fix workflows by enabling the AI engine to compile an application, read error messages, attempt fixes, and recompile automatically.
Optionally, use your own MCP server’s version control capabilities to create and manage pull requests directly from the IDE.

==Feature Summary==
Kai integration to {{Product}} introduces an agentic experience within the IDE, combining a dockable chat interface, inline code completion, and advanced code suggestions that enhance the {{Product}} editing experience.

Kai offers inline autocomplete, which augments standard code completion, while expanded suggestions provide alternative code snippets and refactoring options. It also offers Ghost mode, which lets users add code directly in the editor and accept it with a single keystroke.
Use the Chat window to request coding assistance on creating UI forms, modifying source code, compiling projects, analyzing compiler output, or automatically resolving errors and warnings.

With model selection available per project or even per prompt, Kai offers a highly customizable AI experience, bringing powerful automation and intelligent assistance directly into the RAD Studio development lifecycle.

See all Kai features in more detail in the [[Kai Features]] page.

==See Also==
* [[Kai Installation]]
* [[Kai Features]]
* [[Kai Licensing]]
* [[Kai Options|Kai Settings]]

Fundamental Syntactic Elements (Delphi)

Yukikoi — Tue, 08 Sep 2026 09:22:19 GMT

(Add comment) Removed nonfunctional style attributes and bold markup (''') that caused frequent formatting errors during localization. No change to the rendered output.

Show changes

Windows SDK Installation

L10nBot — Mon, 07 Sep 2026 14:19:18 GMT

updated language links

← Older revision		Revision as of 14:19, 7 September 2026
Line 58:		Line 58:
	* [[SDK Manager]]		* [[SDK Manager]]

			[[de:Installation des Windows SDK]]
	[[ja:Windows SDK のインストール]]		[[ja:Windows SDK のインストール]]

WebBroker Sessions Management

L10nBot — Mon, 07 Sep 2026 14:19:17 GMT

updated language links

← Older revision		Revision as of 14:19, 7 September 2026
Line 112:		Line 112:
	[[Category:C++]]		[[Category:C++]]

			[[de:WebBroker-Sitzungsverwaltung]]
	[[ja:WebBroker セッション管理]]		[[ja:WebBroker セッション管理]]

Structure View (Options)

L10nBot — Mon, 07 Sep 2026 14:19:16 GMT

updated language links

← Older revision		Revision as of 14:19, 7 September 2026
Line 61:		Line 61:
	* [[Object Inspector (Options)]]		* [[Object Inspector (Options)]]

			[[de:Strukturansicht (Optionen)]]
	[[ja:構造ビュー（オプション）]]		[[ja:構造ビュー（オプション）]]

Using DelphiLSP Code Insight with Other Editors

L10nBot — Mon, 07 Sep 2026 14:19:15 GMT

updated language links

← Older revision		Revision as of 14:19, 7 September 2026
Line 151:		Line 151:
	[[Category:Olympus]]		[[Category:Olympus]]

			[[de:Verwenden von DelphiLSP Code Insight mit anderen Editoren]]
	[[ja:他のエディタで DelphiLSP コード支援を使用する]]		[[ja:他のエディタで DelphiLSP コード支援を使用する]]

Finite float

L10nBot — Mon, 07 Sep 2026 14:19:14 GMT

updated language links

← Older revision		Revision as of 14:19, 7 September 2026
Line 35:		Line 35:
	[[Category:{{Delphi}}]]		[[Category:{{Delphi}}]]

			[[de:Begrenzte Gleitkommazahlen]]
	[[ja:有限浮動小数点]]		[[ja:有限浮動小数点]]

Direct2D APIs

L10nBot — Mon, 07 Sep 2026 14:19:14 GMT

updated language links

← Older revision		Revision as of 14:19, 7 September 2026
Line 102:		Line 102:
	* [[Using the Direct2D Canvas]]		* [[Using the Direct2D Canvas]]

			[[de:Direct2D-APIs]]
	[[ja:Direct2D API]]		[[ja:Direct2D API]]

Delphi Upgrade Advisor Wizard

L10nBot — Mon, 07 Sep 2026 14:19:12 GMT

updated language links

← Older revision		Revision as of 14:19, 7 September 2026
Line 118:		Line 118:
	* [[GetIt Package Manager]]		* [[GetIt Package Manager]]

			[[de:Experte "Delphi Upgrade Advisor"]]
	[[ja:Delphi アドバイザウィザードのアップグレード]]		[[ja:Delphi アドバイザウィザードのアップグレード]]

Fundamental Syntactic Elements (Delphi)

Yukikoi — Mon, 07 Sep 2026 12:46:57 GMT

Extended Identifiers

← Older revision		Revision as of 12:46, 7 September 2026
Line 119:		Line 119:
	<source lang="delphi">		<source lang="delphi">
	var TMyType.type		var TMyType.type
	// Using a fully qualified name avoids ambiguity with {{Delphi}} language keyword.		// Using a fully qualified name avoids ambiguity with Delphi language keyword.
	</source>		</source>