Apple is adding Agentic code to Xcode. The company announced on Tuesday that Xcode 26.3 will let developers use Agentic tools like Anthropos’s Claude Agent and OpenAI’s Codex Write in Apple’s official app development suite.
Apple developers can download the Xcode 26.3 release candidate from the developer website today. It will be available at the App Store soon.
This update follows last year’s Xcode 26 release, which added support for ChatGPT and Claude in Apple’s integrated development environment for building apps on iPhone/iPad/Mac/Apple Watch and other Apple devices.
With Agentic coding tools, AI models can leverage more Xcode features to perform tasks and automate more complex processes.
The models will also allow access to Apple’s latest developer documentation, so they can use up-to-date APIs and follow best practices while building. At launch, the agents can help developers explore their projects, understand their structure and metadata, build their projects, run tests, and fix any errors they find.
Apple said it worked closely with Anthropic and OpenAI to design this new experience. The company focused on optimizing token language and tool calls to ensure agents run efficiently in Xcode.
Xcode uses MCP to share its features with the agents and connect them to its tools. This lets Xcode work with any MCP-compatible agent for functions like project discovery, making changes, managing files, previews, code snippets, and accessing the latest documentation.
To try Agentic coding, developers should download the agents they want from Xcode’s settings. They can connect their accounts with the AI providers by signing in or adding an API key. A drop-down menu in the app lets developers choose which model version, such as GPT-5.2 Codex or GPT-5.1 Mini.
Developers can use a prompt box on the left side of the screen to tell the agent what kind of project they want to build or what code changes they want, using natural language. For example, they might ask Xcode to add a feature to their app using one of Apple’s frameworks and describe how it should look and work.
When the Agent starts working, it breaks tasks into smaller steps, making it easy to follow what’s happening and how the code changes. It checks the documentation it needs before coding. Code changes are highlighted, and a project transcript on the side of the screen shows developers what’s happening behind the scenes.
Apple believes this candidness will help new developers who are learning to code. To support this, the company is hosting a Code Along workshop on Thursday on its developer site, where users can watch and learn how to use Agentic coding tools while coding in real-time with their own copy of Xcode.
At the end, the AI agent checks that the code it created works as expected. Based on the test results, the agent can keep working on the project to fix any errors or issues. Apple also noted that asking the agent to plan before writing code can sometimes improve the process since it encourages pre-planning.
If developers are not satisfied with the results, they can easily return their code to its original state at any time since Xcode creates milestones whenever the agent makes a change.
