Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Prosiding Seminar Nasional Sisfotek (Sistem Informasi dan Teknologi Informasi)
Abhy Maulana Fadli
Unknown Affiliation
Computer Science & IT

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Analisis Perbandingan Waktu Respons Model Chatbot AI Generatif (GPT, Gemini, dan DeepSeek) Abhy Maulana Fadli
Prosiding SISFOTEK Vol 9 No 1 (2025): SISFOTEK IX 2025
Publisher : Ikatan Ahli Informatika Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

This study analyzed the response time performance of three generative AI chatbot models GPT-3.5 Turbo, Gemini 2.5 Flash, and DeepSeek Chat V3.1 using a controlled experimental approach. A total of 450 data points were collected through three replicated test sessions involving five categories of questions related to the Laravel framework. Normality and homogeneity tests showed that the data did not meet parametric assumptions, leading to the use of the Friedman test for further analysis. The results indicate a significant difference in response time among the three models, with GPT-3.5 Turbo demonstrating the fastest and most consistent performance, followed by Gemini 2.5 Flash, while DeepSeek Chat V3.1 showed the slowest and most variable responses. Analysis of question types revealed that only GPT-3.5 Turbo exhibited significant performance differences across categories, with Knowledge-Based questions receiving the fastest responses. In contrast, Gemini and DeepSeek maintained stable response times across all question types. These findings highlight that GPT-3.5 Turbo is better suited for real-time applications requiring low latency, while Gemini and DeepSeek offer more consistent performance across varying input complexities. Future research may include newer model versions, broader domains of questions, and evaluation of answer quality for more comprehensive assessment.
Co-Authors