With Xcode 26.3, Apple adds support for agent-based coding, allowing developers to use tools like Anthropic’s Claude Agent and OpenAI’s Codex right in Xcode to build apps.
Agent coding will allow Xcode to complete more complex application development tasks autonomously. Claude, ChatGPT, and other AI models have been available for use in Xcode since Apple added intelligent features to Xcode 26, but until now the AI was limited and unable to act on its own. This will change with the ability to use an AI coding assistant.
AI models have access to more Xcode features to work on a project goal, and Apple has worked directly with Anthropic and OpenAI to configure their agents for use in Xcode. Agents can create new files, explore the project structure in Xcode, build a project directly and run tests, take screenshots to double-check work, and access Apple’s full developer documentation designed for AI agents.
Adding an agent to Xcode can be done with a single click in Xcode settings, and agents can be updated automatically as AI companies release updates. Developers will need to create an Anthropic or OpenAI account to use these coding tools in Xcode and pay fees based on API usage.
Apple says its goal was to make Claude Agent and Codex work efficiently and with reduced token usage. It’s easy to switch between agents in the same project, giving developers the flexibility to choose the agent best suited for a particular task.
While Apple worked with OpenAI and Anthropic to integrate Xcode, Xcode 26.3 features can be used with any agent or tool that uses the open standard Model Context protocol. Apple publishes documentation to enable developers to configure and connect MCP agents to Xcode.
Using natural language commands, developers are able to instruct AI agents to complete a project, such as adding a new feature to an app. Xcode then works with the agent to break down the instructions into small tasks, and the agent is able to work on its own from there. The procedure works like this:
- A developer asks an embedded agent to add a new feature to an application.
- The agent looks at the current project to see how it is organized.
- The agent checks all relevant documentation, looks at code snippets, code samples, and the latest API.
- The agent starts working on the project and continuously adds code.
- The agent builds the project and then uses Xcode to verify its work.
- If errors or warnings are encountered, the agent continues to work until all problems are resolved. He is able to access the build logs and revise until the project is perfect.
- At the end, the agent provides a summary of everything that happened so that developers have a clear overview of the implementation.
In the project sidebar, developers can watch what the agent is doing using the script and click to see where the code is being added to get an overview of what the agent is doing. Developers can always go back to when an agent or model made an adjustment, so there are options to undo unwanted results or try multiple options to introduce a new feature.
Apple says agent-based coding will enable developers to streamline workflows, accelerate changes and bring new ideas to life. Apple also sees it as a learning tool that gives developers the opportunity to learn new ways to build something or implement an API into an app.
“At Apple, our goal is to create tools that put cutting-edge technology directly into the hands of developers so they can build the best apps,” said Susan Prescott, Apple’s vice president of worldwide developer relations. “Agent coding increases productivity and creativity, and simplifies the development workflow, so developers can focus on innovation.”
The Xcode 26.3 release candidate is available to developers starting today, with a launch likely to follow in the next week or so.