The increasing use of generative artificial intelligence (AI) in language learning has created new opportunities for translation and language-material development. However, limited research has examined students’ perceptions of using ChatGPT, a generative AI-based conversational assistant developed by OpenAI, in product-oriented activities such as developing a Classroom Language Dictionary. This study investigated postgraduate students’ perceptions of using ChatGPT for translation and Classroom Language Dictionary development. A descriptive mixed-methods design was employed with nine English Language Education postgraduate students at a private university in Indonesia. Data were collected through a self-administered questionnaire consisting of nine four-point Likert-scale items and one open-ended question. Quantitative data were analyzed using frequencies, percentages, and means while qualitative responses were analyzed thematically. The findings revealed strongly positive perceptions across the three dimensions of Robbins and Judge: perceiver, target, and situation. Students perceived AI as engaging and supportive of their translation development, considered it relevant to their future teaching needs, and viewed the learning environment as conducive to guided and critical AI use. The open-ended responses showed that AI supported translation and language refinement, vocabulary and contextual language development, content generation and organization, and feedback, critical evaluation, and adaptation. These findings suggest that AI can function effectively as a learning assistant when integrated with teacher guidance and critical evaluation.