Comparative Analysis of GenAI Models in Problem Solving on the LeetCode Platform

Aslan Zhaksylykov, Zhanat Kunapianovna Nurbekova, Ainur Sagimbayeva, Shirinkyz Tleubergenovna Shekerbekova

2025

Closed access Metadata (CSL-JSON)
DOI 10.1109/ubmk67458.2025.11206938
Data confidence 85 high
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

This paper compares the performance of different AI models (GPT-4o, Gemini 2.0, GPT o3-mini, Grok, and LLAMA) in solving programming problems on the Leetcode platform. The study assessed the accuracy, speed, and ability of the models to handle complex cases. The study identified the strengths and weaknesses of each model, as well as issues with the accuracy of calculations. The results of the study demonstrate the importance and features of using GenAI in teaching programming, but also indicate the need to develop critical thinking and student independence. Recommendations for optimizing educational tasks taking into account the use of GenAI are proposed.