DeepSeek-R1: Reliability for research and education – A comparative study with Claude-3.5-Sonnet and GPT-4o
Journal of Clinical and Basic Research · 01.10.2025 · 9 (3) · 1-5
DOI: 10.66224/jcbr.9.3.1 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Clinical and Basic Research, online erschienen am 01.10.2025. Verfasst von Reza Rakhshi und 4 weiteren Autor:innen. Die Arbeit nennt 12 Literaturangaben.
Abstract (Originalsprache)
Background
Large language models (LLMs) like Claude-3.5-Sonnet and GPT-4o are widely used in research and education but are limited by high costs and proprietary restrictions. DeepSeek-R1, an open-source LLM developed by DeepSeek-AI, leverages a Mixture-of-Experts (MoE) architecture and multi-stage training to offer a cost-effective alternative. This study evaluates DeepSeek-R1’s reliability for academic and clinical applications compared to Claude-3.5-Sonnet and GPT-4o, focusing on performance, cost efficiency, and limitations such as censorship and data privacy.
Methods
A mixed-methods approach was employed, including benchmark evaluations across MATH-500 (mathematics), HumanEval (programming), MMLU (general knowledge), and MedQA (medical reasoning). A prospective user study with 112 Iranian medical researchers assessed diagnostic accuracy on 50 standardized medical cases across specialties (internal medicine, pediatrics, psychiatry). Performance was measured as mean accuracy ± SD, with paired t-tests (p
Zitieren
APA
Rakhshi, R., Khosravi, T., Rahimzadeh, A., Mohsenipour, M., & Oladnabi, M. (2025). DeepSeek-R1: Reliability for research and education – A comparative study with Claude-3.5-Sonnet and GPT-4o. Journal of Clinical and Basic Research, 9(3), 1-5. https://doi.org/10.66224/jcbr.9.3.1
Vancouver
Rakhshi R, Khosravi T, Rahimzadeh A, Mohsenipour M, Oladnabi M. DeepSeek-R1: Reliability for research and education – A comparative study with Claude-3.5-Sonnet and GPT-4o. Journal of Clinical and Basic Research. 2025;9(3):1-5. doi:10.66224/jcbr.9.3.1
BibTeX
@article{Rakhshi2025,
author = {Rakhshi, Reza and Khosravi, Teymoor and Rahimzadeh, Arian and Mohsenipour, Mohadeseh and Oladnabi, Morteza},
title = {DeepSeek-R1: Reliability for research and education – A comparative study with Claude-3.5-Sonnet and GPT-4o},
journal = {Journal of Clinical and Basic Research},
year = {2025},
volume = {9},
number = {3},
pages = {1--5},
doi = {10.66224/jcbr.9.3.1}
}
EUVIMED