Developer experience (DX) : a comparative analysis of Streamlit, Chainlit, and Gradio
Naeem, Muhammad Zain (2025)
2025
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025121837825
https://urn.fi/URN:NBN:fi:amk-2025121837825
Tiivistelmä
The rapid rise of chatbot applications has created a high demand for efficient conversational user interfaces. However, selecting the right framework to develop an interface is often based on subjective marketing rather than objective data. The perceptions and feelings of the developer are mostly ignored. The thesis quantitatively evaluated Streamlit, Gradio, and Chainlit based on Developer Experience (DX) to address this issue.
In the research, three functionally identical chatbots were developed, implementing message streaming, state management, and multimodality. Static code analysis was performed on the resulting source code which provided the number of operators and operands, termed as tokens altogether, present in the code logics. These tokens calculated Halstead’s Complexity Measures and McCabe’s Cyclomatic Complexity, which represented implementation effort and structural complexity of the code. An analysis was performed to equate the calculated values to the identified DX framework.
Streamlit records the highest difficulty and cognitive load due to its script-based architecture. Gradio proves to be the most developer friendly framework, requiring nearly 66% less mental effort. Chainlit demonstrates the lowest structural complexity, making it suitable for scalable applications. The study concludes that Gradio offers the superior Developer Experience for standard chatbots.
In the research, three functionally identical chatbots were developed, implementing message streaming, state management, and multimodality. Static code analysis was performed on the resulting source code which provided the number of operators and operands, termed as tokens altogether, present in the code logics. These tokens calculated Halstead’s Complexity Measures and McCabe’s Cyclomatic Complexity, which represented implementation effort and structural complexity of the code. An analysis was performed to equate the calculated values to the identified DX framework.
Streamlit records the highest difficulty and cognitive load due to its script-based architecture. Gradio proves to be the most developer friendly framework, requiring nearly 66% less mental effort. Chainlit demonstrates the lowest structural complexity, making it suitable for scalable applications. The study concludes that Gradio offers the superior Developer Experience for standard chatbots.