This paper presents a comprehensive review of Artificial General Intelligence (AGI) and Agentic AI, examining their definitions, evolution, technological foundations, current capabilities, and future trajectories. We analyze AGI’s theoretical underpinnings, contrasting it with narrow AI and artificial superintelligence (ASI), and highlight three critical dimensions of AGI development: (1) technical architectures bridging narrow AI to general intelligence, (2) transformative applications across sectors such as finance, business, and research, and (3) emerging ethical and workforce challenges. Agentic AI is also explored in depth, with a focus on its architectural requirements, computational demands, and learning mechanisms.We review specialized frameworks like OpenAI’s AGI classification and Agentic AI platforms such as Vectara-agentic and CrewAI, along with enabling infrastructure including NVIDIA’s AI hardware and large-scale cloud systems like OpenAI’s proposed "Stargate." Our comparative analysis reveals that Agentic AI is already being deployed in areas such as legal services, DevOps, and enterprise automation, whereas AGI remains in the research stage with ongoing debates about feasibility and timelines.Market data shows accelerating AGI growth (36.9% CAGR through 2031) and strong adoption in financial services (38% of AI investments projected by 2028). Despite this momentum, significant gaps persist in areas such as evaluation metrics, environmental impact, and governance. We propose a framework for responsible AGI development that balances innovation with ethical standards, emphasizing standardized benchmarks and workforce transition strategies.This is a pure review paper; all findings are derived from cited literature. It contributes to ongoing discourse by synthesizing fragmented research into actionable insights for practitioners and policymakers navigating the AGI and Agentic AI landscape.