Agentic coding comes to Apple’s Xcode with agents from Anthropic and OpenAI

Apple brings agentic encryption to Xcode. The company has Tuesday announced the release of Xcode 26.3, which allows developers to use agentic tools, including those from Anthropic Claude Agent and OpenAI’s Codex, directly in Apple’s official app development suite.

The Xcode 26.3 Release Candidate is available today to all Apple developers via the developer website and will appear in the App Store a little later.

This latest update follows last year’s Xcode 26 release, which for the first time introduced support for ChatGPT and Claude within Apple’s integrated development environment (IDE) used by those building apps for iPhone, iPad, Mac, Apple Watch, and Apple’s other hardware platforms.

By integrating agentic coding tools, AI models can leverage more features of Xcode to perform their tasks and perform more complex automation.

The models will also have access to Apple’s current developer documentation to ensure they are using the latest APIs and following best practices during build.

At launch, the agents can help developers explore their project, understand its structure and metadata, then build the project and run tests to see if there are any bugs and fix them if so.

In preparation for this launch, Apple says it has worked closely with both Anthropic and OpenAI to design the new experience. Specifically, the company said it has done a lot of work to optimize token usage and tool invocations so that the agents could work efficiently in Xcode.

Xcode uses MCP (Model Context Protocol) to expose its capabilities to the agents and connect them to its tools. That means Xcode can now work with any third-party MCP-compliant agent for things like project discovery, changes, file management, previews and snippets, and access to the latest documentation.

Developers who want to try the agentic coding feature must first download the agents they want to use from the Xcode settings. They can also connect their accounts to the AI ​​providers by logging in or adding their API key. A drop-down menu within the app allows developers to choose which version of the model they want to use (e.g. GPT-5.2-codex vs. GPT-5.1-mini).

In a prompt box on the left side of the screen, developers can tell the agent what kind of project they want to build or what kind of project they want to create in the code they want to create using natural language commands. For example, they can instruct Xcode to add a feature to their app that uses one of Apple’s provided frameworks, and how it should look and function.

When the agent starts working, it breaks down the tasks into smaller steps so you can easily see what’s happening and how the code is changing. It will also look for the documentation it needs before it starts coding. The changes are visually highlighted in the code, and the project transcript on the side of the screen allows developers to see what’s happening under the hood.

This transparency could especially help new developers learning to code, Apple believes. The company hosts for this purpose a “code-along” workshop on Thursday on the 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 of its process, the AI ​​agent verifies that the code it created works as expected. Armed with the results of its testing on this front, the agent can iterate on the project further if necessary to fix errors or other issues. (Apple noted that asking the agent to think about its plans before writing code can sometimes improve the process because it forces the agent to plan in advance.)

And if developers aren’t happy with the results, they can easily revert their code back to the original code at any time, because Xcode creates milestones every time the agent makes a change.