本研究探讨大语言模型(LLMs)在大学英语教学中的应用潜力，旨在为语言教学提供研究借鉴及实践启发。研究结合《新时代大学进阶英语综合教程 Book 2》教学案例展示多种 LLMs 赋能词汇、语法、阅读、口语、写作跨场景的应用。研究发现，LLMs 可实现主题词汇动态分类与多模态可视化呈现，构建语义场以强化学生词汇关联学习；依托错误生成与对比分析功能，促进语法规则精讲与分层训练；通过情感尺度分割与隐性叙事追踪，助力学生批判性思维与同理心能力提升；结合语音识别技术实现口语表达效果客观化评估；通过生成式修正完善写作文本连贯性。LLMs 通过深度语义理解与生成能力，可以显著优化语言知识教学与技能培养的精准性与效率。

This paper explores the potential applications of Large Language Models (LLMs) in college English teaching and aims to provide research references and practical insights. Five cross-scenarios of LLMs are demonstrated, namely, the optimization of vocabulary, grammar, reading, speaking, and writing teaching links, using Intermediate College English for New Era (Book 2) as a case study. It is found that LLMs enable dynamic categorization of thematic vocabulary and multimodal visualization to construct semantic fields, thereby reinforcing students’ associative vocabulary learning; through error-generation mechanisms and comparative text analysis, they facilitate targeted instruction of grammatical rules with hierarchical training adaptations; by implementing sentiment scale segmentation and implicit narrative tracking, they foster the development of critical thinking and empathetic reasoning competencies; when integrated with speech recognition technology, they permit objective evaluation of oral proficiency metrics; their generative revision mechanisms effectively enhance textual coherence in writing outputs. Leveraging deep semantic comprehension and generative capacities, LLMs significantly optimize the precision and efficiency of language knowledge acquisition and skill development in educational contexts.