With the rapid development of large language models in text generation, semantic understanding, and translation assistance, their potential application in literary translation has become increasingly prominent.Gong'an literature(公案文學) is a distinctive literary genre unique to ancient China and a concentrated literary embodiment of the spirit of the rule of law in traditional Chinese society.As a literary form that combines narrative, judicial, cultural, and historical dimensions, the translation of gong'an literature involves not only the cross-linguistic transfer of plot, legal terminology, and judicial reasoning, but also the effective transmission of ancient Chinese judicial culture, ethical values, and the spirit of the rule of law.Compared with general literary texts, the translation of gong'an literature faces multiple challenges, including the need to balance plot development with legal expression, the intertwining of historical context with institutional meaning, and the complexity of character discourse and adjudicative logic.Against this background, this article takes "the theoretical logic and practical pathways of empowering the translation of gong'an literature through large language models" as its central focus, exploring the practical value, scope of applicability, and optimization strategies of employing large language models in the translation of gong'an literature.This article first examines the textual features of gong'an literature and analyzes the advantages of large language models in terminology recognition, discourse restructuring, contextual supplementation, multi-version generation, and translation assistance. It argues that the theoretical logic underlying the empowerment of gong'an literature translation by large language models can be understood in four dimensions: first, a language generation logic based on contextual prediction and semantic association; second, a knowledge-enhancement logic oriented toward the interpretation of ancient legal culture; third, a human–machine collaborative logic that emphasizes the translator's leading role combined with model assistance; and fourth, a logic of meaning transformation serving the international dissemination of Chinese legal culture. At the same time, this article also points out that large language models still suffer from such problems as misjudgment of historical context, misreading of legal culture, smoothing of narrative style, weakening of culturally loaded information, and hallucinated generation.They therefore remain unable to replace the central role of human translators in cultural discernment, institutional interpretation, and the reproduction of literary style.On this basis, the article proposes a human–machine collaborative pathway for the translation of gong'an literature.By constructing an integrated support system combining a corpus, a terminology database, and a knowledge base, it advocates a multi-layered translation process consisting of model-generated draft translation, human revision, expert review, and dissemination feedback, so as to achieve a balance among translation efficiency, academic accuracy, and cultural communication.This article argues that the value of large language models in the translation of gong'an literature lies not in replacing human translators, but in serving as intelligent auxiliary tools that enhance the organizational capacity, interpretive depth, and communicative effectiveness of translating highly complex cultural texts.Exploration of this issue not only helps expand the interdisciplinary horizon between intelligent translation studies and research on the translation of ancient Chinese literature, but also provides new methodological references for the international dissemination of ancient Chinese legal culture.