Xojo, Inc. has released Xojo 2026 Release 3 (2026r3), introducing Agentic Mode to let AI coding agents work directly with projects inside its cross-platform development environment. The feature allows agents to make code changes, trigger project execution and inspect compiler feedback, while the IDE automatically reloads changes made externally. The release also updates Xojo’s built-in AI assistant, Jade, and adds improvements across desktop, web, iOS and Android development, reflecting the growing move toward AI agents that participate in software development workflows rather than simply generate code snippets.
Xojo Introduces Agentic Mode for AI Coding Workflows
Xojo has introduced Agentic Mode in its latest development environment release, giving AI coding agents a more direct role in building and testing applications within the Xojo integrated development environment (IDE).
The feature, included in Xojo 2026 Release 3, allows the IDE to reload project changes made by external tools and lets an agent run a project to obtain compilation feedback. Instead of generating code that developers must manually transfer, integrate and test, an agent can work against the project while Xojo provides feedback from its compiler.
The approach targets a common limitation in AI-assisted programming. Code-generation tools can produce functions, classes and application components quickly, but generated code still needs to fit the project’s structure, dependencies and language rules. Developers may otherwise spend time moving code between tools, refreshing files and diagnosing errors that emerge only after compilation.
Agentic Mode is designed to shorten that loop. Xojo says developers can enable the feature through File > Agentic Mode, allowing AI agents to modify project files and have those changes reflected in the IDE. Agents can then run the project and use compiler errors to guide subsequent changes.
The developer remains in control of the workflow, according to the company. However, the announcement does not identify specific supported coding agents, explain their connection and permission mechanisms, or provide benchmarks showing how much development time the feature saves.
Project Reloading Connects the IDE With External Tools
Alongside Agentic Mode, Xojo has expanded its project-reloading capabilities. Developers can reload an entire project, individual project items or only items that have changed on disk.
This is important for workflows that combine an IDE with external coding agents, scripts and version-control tools. When multiple tools modify the same project, an IDE must recognize those changes without forcing developers to repeatedly close and reopen files or risk working with stale versions.
Selective reloading can help developers keep the project synchronized while limiting unnecessary disruption. It also supports a broader development pattern in which the IDE becomes one component of a workflow involving AI agents, source-control systems and automated testing.
The practical value will depend on how reliably the IDE detects changes, resolves conflicts and handles concurrent edits. Xojo’s announcement describes the new reloading options but does not detail conflict-resolution behavior or integration with particular agent frameworks.
Jade Assistant Receives Reliability Improvements
Xojo has also updated Jade, its built-in AI assistant. The latest changes improve copying and pasting between the assistant and text, handling of special characters on Windows, and support for longer-running requests.
These improvements address the less visible friction that can undermine AI-assisted development. Coding assistants must preserve punctuation, escaping and other syntax-sensitive characters when moving code between interfaces. Long-running requests also require an interface that can manage responses without interrupting the developer’s workflow.
Jade’s updates complement Agentic Mode, although the announcement does not specify whether Jade itself can operate as an autonomous agent through the new feature or how its capabilities compare with third-party coding agents.
The distinction matters because AI assistants and coding agents are not interchangeable. An assistant may explain code or suggest edits, while an agent can execute a sequence of actions, inspect results and continue working toward a goal. Agentic Mode focuses on enabling that more iterative interaction with Xojo projects.
Desktop, Web and Mobile Development Get Updates
The release also includes a broad set of changes across Xojo’s supported development targets.
For desktop applications, the new DesktopTable control adds another option for building data-oriented interfaces. Other changes include symbols in menu items, system images available across targets, and improvements for Windows applications using WinUI and for Linux apps.
Web developers receive performance improvements for populating large ListBox controls through ListBox.AddRow. The release also adds WebCanvas.DrawControlInLayoutEditor for design-time previews of custom controls and WebRectangle.Pressed for responding to presses on web interface elements.
Mobile development receives updates across iOS and Android. These include dynamic type support, popup-menu separators and document pickers. iOS also gains a document viewer, printer picker and MobileSearchField, while Android adds toolbar-button badges and picture-mask support.
These features address a different but complementary challenge: maintaining a consistent development experience across platforms while accommodating each operating system’s interface conventions and capabilities. Xojo’s cross-platform model aims to let developers share more of their application logic while still accessing platform-specific functionality.
The release also updates DBKit v2 with support for DatabaseConnection, related-table display and lazy loading. These additions target applications that work with connected data and larger sets of related records.
Agentic Development Raises New Workflow Questions
Xojo’s release comes as software development tools increasingly move beyond conversational code suggestions toward agents that can edit files, execute code and respond to tool feedback. The key distinction is the integration of AI into the development loop, rather than treating code generation as a separate step.
Giving agents access to project execution and compiler feedback can make it easier to identify basic errors and iterate on changes. But compilation is only one measure of software correctness. A project can compile successfully while still containing logic defects, security vulnerabilities, poor user experiences or failures that appear only at runtime.
Developers will therefore still need to review agent-generated changes, test behavior and apply appropriate security and release controls. Agent permissions, auditability and recovery from unwanted edits also become important when an automated system can modify project files and run applications.
Xojo 2026r3 establishes a more direct path for AI agents to work inside its IDE. Whether that translates into substantial productivity gains will depend on compatibility with coding agents, reliability across real projects and how well developers can supervise the resulting changes.
Market Landscape
AI-assisted software development is progressing from code completion and chat-based suggestions toward agents that can edit files, run tools and iterate on errors. IDE vendors are increasingly competing on how tightly these capabilities integrate with existing development workflows.
Xojo’s Agentic Mode connects AI agents with project reloading, execution and compiler feedback inside its cross-platform IDE. The release also adds platform-specific controls and updates for desktop, web and mobile developers.
The central test will be whether this integration reduces manual handoffs without weakening developer oversight. Compilation feedback can help agents correct syntax and build errors, but testing, security review and human validation remain necessary for production software.
Top Insights
- AI agents work directly with projects: Agentic Mode lets agents modify Xojo projects, run applications and use compiler feedback to guide further changes.
- Project synchronization improves: Developers can reload complete projects, individual items or only files that changed on disk.
- Jade becomes more reliable: Updates improve text transfer, Windows special-character handling and longer-running AI requests.
- Cross-platform development expands: DesktopTable, web performance updates, mobile document tools and dynamic type support extend the IDE’s capabilities.
- Human oversight remains necessary: Compilation feedback can catch build errors, but it does not guarantee correct behavior, security or production readiness.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
