Explainable Artificial Intelligence (XAI) refers to a class of methods and tools that make the decision-making processes of AI systems transparent, understandable, and accountable to human users, particularly in high-stakes applications such as healthcare, finance, autonomous systems, and cybersecurity where opaque models can hinder trust, safety, and compliance. With growing ethical and regulatory concerns around black-box AI models, XAI has become essential for ensuring interpretability, fairness, and responsible AI deployment. This paper presents a comprehensive literature review on XAI by first establishing its conceptual foundation, including definitions, explanation types, and the needs of various stakeholders. It then reviews a wide array of XAI techniques, distinguishing between model-specific and model-agnostic methods, and highlights visualization tools, surrogate models, and rule-based explanations. The study further analyzes prominent XAI libraries and platforms such as InterpretML, AIX360, Captum, and AWS Clarify, using evaluation criteria like fidelity, stability, complexity, and generalizability. Real-world applications across critical domains are discussed to demonstrate the value of XAI in enhancing trust and decision support. Finally, the paper identifies key challenges such as trade-offs between accuracy and interpretability, lack of standards, and explainability in deep models, while proposing future research directions involving causal inference, federated AI, human-centric design, and transparency in large language models and reinforcement learning systems.