Many AI Agents use Large Language Models (LLMs) for language.

But agents often do more:

  • Planning: They figure out how to do things step by step.
  • Memory: They remember what they've done.
  • Tool Use: They use other programs and tools.
  • Interaction: Some can see and act in the real world.

So, while LLMs are a key part, calling an AI Agent just an LLM wrapper misses the other important pieces that make them work.