LLM Agents, or Language Model Agents, are advanced systems that enhance the capabilities of traditional conversational LLMs by integrating components like memory, tools, and planning. These agents can interact with their environment, use external tools (e.g., calculators, web search), and plan actions autonomously. Key components include short-term and long-term memory systems, tool use (e.g., function calling, Toolformer), and planning techniques like ReAct and Reflexion. Multi-Agent frameworks, such as Generative Agents and CAMEL, enable collaboration between specialized agents, each with unique tools and memory systems, to solve complex tasks. These advancements are driving the evolution of LLMs into more autonomous and capable systems.