DeepSeek Review: See How It Performs in Programming Skills Assessment

In a world where artificial intelligence continues to transform our daily lives, the emergence of new players often generates a lot of buzz. Recently, DeepSeek, an open-source Chinese AI chatbot, has attracted attention. By outperforming well-established benchmarks in programming tests, it raises questions about the future of AI tools. How does its performance differ from that of industry giants, and what implications could this have for developers and programming education? Let’s explore DeepSeek’s capabilities while examining its potential and limitations.

DeepSeek’s Performance in Coding Tests

In an article on ZDNET, David Gewirtz based his evaluation of DeepSeek on a set of well-defined programming tests. The appeal of this approach lies in the ability to compare DeepSeek’s performance with that of other models, such as those developed by OpenAI, Google, and Microsoft. In 2025, as competition intensifies, these evaluations can help draw informed conclusions about the current AI landscape.

The first test: creating a WordPress plugin

The inaugural test asked DeepSeek to create a simple WordPress plugin. The goal was to manage a list of names, ensuring that duplicates were separated. This type of project is common in the development world, especially for those training with platforms like Codecademy and OpenClassrooms. To Gewirtz’s surprise, DeepSeek responded successfully, proving its ability to understand the project’s requirements. This result is significant: only half of the models tested succeeded in this task. The second test: rewriting a validation function The next challenge required DeepSeek to modify a validation routine to accept not only dollar amounts, but also cents. Here, DeepSeek demonstrated impressive skills, although its code could have been optimized. It demonstrated a solid understanding of validation, even if some experts consider it too verbose. This contrast between efficiency and clarity is often a dilemma for those taking coding courses on platforms like Khan Academy or Udemy. Comparative Analysis: DeepSeek vs. CompetitorsTo understand DeepSeek’s position in the market, it’s helpful to compare it to other giants in the field. DeepSeek’s ability to pass certain tests with flying colors suggests it could potentially outperform better-known models. This raises questions about the future of learning tools. Will they all specialize like DeepSeek, or will they try to improve aspects like the user interface or speed? In this section, let’s examine a comparison table based on the test results of different AIs.

AI

Test 1: WordPress Plugin Test 2: Amount Validation Test 3: Bug Detection Test 4: Script WritingDeepSeek

Passed

Passed	Failed	ChatGPT	Passed	Passed
Passed	Gemini	Passed	Failed	Passed
Passed	Copilot	Passed	Passed	Failed
Passed	Test Three: Bug Detection	The third challenge involved solving a complex bug encountered by the article’s author. This required a deep understanding of specific API calls and the ability to look beyond error messages. DeepSeek was able to detect the issue, earning it extra points for problem-solving. In an environment where bugs can be costly in terms of time and resources, skills like these are invaluable. Developers using training platforms like Le Wagon or Simplon	often face such scenarios.	Fourth Test: Writing a Complex Script
The final test presented a real challenge for DeepSeek. It involved creating an engaging script in multiple environments, including AppleScript and Keyboard Maestro. While DeepSeek had its shortcomings, this highlights a crucial point for those learning to code: the complexity of tool interaction.	Lessons to be learned from the DeepSeek experience	Each interaction with DeepSeek reveals not only its strengths, but also its weaknesses. While AI’s performance in simple coding tasks can be impressive, errors in more complex scenarios show there is still a long way to go. This opens a discussion about the future of programming education. Developer requirements are evolving, and so are the tools they use.	DeepSeek and the Future of Technology Education	The rise of technologies like DeepSeek has the potential to transform our approach to technology education. Online platforms such as OpenClassrooms, Codecademy, and Coursera could benefit from integrating AI to further enhance and personalize the learning journey. This could offer new tools for assessment, explanation, and support based on individual performance.

Integration of AI into learning systems

Personalization of learning paths Help with specific problem solving Instant feedback on practical exercises Conclusion on the promises and challenges of DeepSeek While DeepSeek has shown promising results in some tests, it remains clear that it must continue to evolve to meet increasingly sophisticated coding needs. It remains to be seen how this technology will influence the programming and education landscape. As competition becomes increasingly fierce, tools like DeepSeek could represent a turning point in how we learn to code and solve problems. What does the future hold? Could the future see an even greater integration of training platforms with innovative AI?