How to Implement iOS Screen Sharing with ReplayKit: A Step‑by‑Step Guide

Background

iOS applications can implement screen sharing to allow users to broadcast their screen content as video during video calls or interactive live streams, enhancing interaction and communication efficiency. This feature is valuable in video conferencing, online education, and game live streaming scenarios.

Screen sharing can be achieved via two approaches: in‑app sharing and cross‑app sharing, with Apple's ReplayKit framework providing the necessary APIs.

ReplayKit Introduction

ReplayKit, introduced in iOS 9, enables developers to record and live‑stream screen content and audio, including microphone input. It offers simple APIs for screen capture, supports extensions for real‑time data processing, and has evolved through iOS versions to add features such as global recording, direct launch, and Loop Buffer for continuous short‑duration recording.

iOS 9: Initial release, supports in‑app screen and audio recording, generates MP4 files.

iOS 10: Added real‑time video data access for live streaming via extensions.

iOS 11: Supported global system‑wide recording via Control Center.

iOS 12: Introduced direct launch with RFSystemBroadcastPickerView , simplifying user workflow.

iOS 15: Added Loop Buffer for up to 15 seconds of continuous recording.

Cross‑App Sharing Process

The overall architecture for iOS video‑conference screen sharing using cross‑app sharing is illustrated below.

The sharing workflow includes:

User prepares the app.

User launches ReplayKit from Control Center.

ReplayKit Extension starts receiving audio‑video streams and pushes them to the server.

User stops recording from Control Center, ending the process.

Implementation Details

Key steps for developers:

Ensure devices run iOS 12.0+ and have sufficient performance (recommended iPhone X or later).

Create a Broadcast Upload Extension target in Xcode.

Configure an App Group to share data between the main app and the extension.

In the extension’s SampleHandler , implement broadcastStartedWithSetupInfo , broadcastPaused , broadcastResumed , and broadcastFinished to manage the sharing lifecycle. Process captured buffers in processSampleBuffer:withType: , handling video frames ( RPSampleBufferTypeVideo ) and audio frames ( RPSampleBufferTypeAudioApp , RPSampleBufferTypeAudioMic ) and forward them using a video‑cloud RTC SDK.

Provide UI controls in the app to start/stop sharing, typically using RPSystemBroadcastPickerView , or allow users to invoke sharing from Control Center.

Conclusion

ReplayKit has progressed from basic in‑app recording to system‑wide capture with advanced features like Loop Buffer, but its APIs remain complex and privacy‑focused, making implementation challenging. Developers must handle iOS version compatibility, memory limits (extension processes must stay under 50 MB), and inter‑process communication (via sockets, App Group NSUserDefaults , or CFNotificationCenter ). Improper termination can cause subsequent sharing attempts to fail, often requiring a device reboot.