This paper presents a systematic literature review of the application of large language models (LLMs) in cybersecurity, analyzing over 180 works across 25 LLMs and more than 10 downstream scenarios. The study addresses three key research questions: how to construct cybersecurity-oriented LLMs, what are the potential applications of LLMs in cybersecurity, and what are the existing challenges and future research directions in this area. The review highlights the potential of LLMs in enhancing cybersecurity practices, including tasks such as vulnerability detection, secure code generation, program repair, binary analysis, threat intelligence, and offensive security. It also discusses the challenges of applying LLMs in cybersecurity, such as inherent vulnerabilities and the need for further research. The paper provides a comprehensive overview of existing cybersecurity LLMs, their applications, and the challenges they face. It also offers practical guidance for researchers and practitioners in applying LLMs to cybersecurity. The study concludes that LLMs have significant potential in cybersecurity, but further research is needed to address the challenges and improve their effectiveness. The paper also provides a list of practical guides on LLMs for cybersecurity at https://github.com/tmylla/AwesomeLLM4Cybersecurity.This paper presents a systematic literature review of the application of large language models (LLMs) in cybersecurity, analyzing over 180 works across 25 LLMs and more than 10 downstream scenarios. The study addresses three key research questions: how to construct cybersecurity-oriented LLMs, what are the potential applications of LLMs in cybersecurity, and what are the existing challenges and future research directions in this area. The review highlights the potential of LLMs in enhancing cybersecurity practices, including tasks such as vulnerability detection, secure code generation, program repair, binary analysis, threat intelligence, and offensive security. It also discusses the challenges of applying LLMs in cybersecurity, such as inherent vulnerabilities and the need for further research. The paper provides a comprehensive overview of existing cybersecurity LLMs, their applications, and the challenges they face. It also offers practical guidance for researchers and practitioners in applying LLMs to cybersecurity. The study concludes that LLMs have significant potential in cybersecurity, but further research is needed to address the challenges and improve their effectiveness. The paper also provides a list of practical guides on LLMs for cybersecurity at https://github.com/tmylla/AwesomeLLM4Cybersecurity.