This study aims to analyze the quality of Arabic–Indonesian translation generated by DeepSeek AI using the book Al-‘Arabiyyah li al-Nāsyi’īn as the object of study. The research employs a descriptive qualitative approach with a content analysis design. The data source consists of Arabic narrative texts taken from Al-‘Arabiyyah li al-Nāsyi’īn Volume 1, which were translated into Indonesian using DeepSeek AI. Data collection techniques include observation and note-taking, while data analysis follows three stages: data reduction, data presentation, and conclusion drawing. The analysis is based on Peter Newmark’s semantic translation theory. The findings indicate that DeepSeek AI is generally capable of producing accurate, clear, and acceptable translations. The translated texts successfully convey the main meaning of the source language with minimal ambiguity and distortion. DeepSeek AI demonstrates a tendency toward semantic translation by adjusting sentence structure, diction, and contextual meaning to improve readability and naturalness in the target language. Although some literal translations are still found, they do not significantly affect overall comprehension. Therefore, DeepSeek AI can be considered an effective supporting tool for Arabic–Indonesian translation, particularly for simple narrative texts and Arabic language learning
Copyrights © 2025