Large Language Models (LLMs) have become more popular in recent years and are now used in various areas of software development because of their ability to understand and generate natural language, including source code. With more powerful LLMs available, developers face the challenge of choosing the right one for generating source code. While some studies have looked at tools like DeepSeek or ChatGPT, there is little research on how developers can pick the best model for their needs. It's important to know if a model can generate useful code that meets quality standards and if developers can use the code effectively. This paper presents a way to compare different models by looking at how two models perform. We tested them on 25 programming tasks using Python language. Both models generated good code, but each had few issues. We investigated the functional and non-functional qualities of the code synthesized by the models on a program synthesis benchmark containing these 25 tasks. Overall, our evaluation shows that overall DeepSeek performs better based on this comparison, which is also supported by human reviewers who checked the generated code manually.

Code Generator, DeepSeek, ChatGPT, Developers, Software Development