Kotlin Multiplatform, Flutter and Native iOS/Android: A Performance Comparison

Abstract

This thesis provides an in-depth performance comparison of mobile development approaches, concentrating on Kotlin Multiplatform, Flutter, and native development for iOS and Android. With the growing demand for cross-platform mobile solutions, frameworks like Kotlin Multiplatform and Flutter strive to lower development efforts, accelerate time to market, and improve code maintainability by enabling the sharing of business logic across platforms. However, it is important to consider whether these benefits come at the expense of runtime performance or other aspects such as memory usage, battery efficiency, or user experience, particularly in applications where speed, responsiveness, and system efficiency are vital. Several practical benchmark scenarios were defined and consistently implemented across all platform technologies to address this issue. These scenarios include network operations such as RESTful HTTP GET requests, database interactions for inserting and reading entries, rendering data in user interfaces, performing CPU-intensive arithmetic computations, and accessing native device APIs like accelerometer sensors and GNSS. The tests were conducted under controlled conditions on comparable hardware, and quantitative metrics such as execution time, memory usage, CPU usage, and app size were analyzed, as well as qualitative aspects like ease of integration and lines of code. The findings indicate that native iOS applications consistently provide the best performance and resource efficiency. Flutter delivers strong CPU performance, but is less efficient in terms of memory usage and database access speed. Kotlin Multiplatform offers good overall performance and native UI flexibility, although it involves higher development complexity. Native Android apps provide balanced performance with a low entry barrier, but slightly underperform in critical benchmarks and database handling. Developers, architects, and organizations can gain valuable insights from this study to make informed decisions about mobile technology stacks. Clearly outlining the tradeoffs between performance, development efficiency, and code reuse helps to evaluate whether modern cross-platform frameworks can be viable alternatives to native development in specific project contexts.

Date of Award	2025
Original language	English
Supervisor	Jens Krösche (Supervisor)