What is an AI agent vs a simple LLM call?
LLM & AILevel: FoundationAgent Fundamentals
What is an AI agent vs a simple LLM call?
Hints
- Identify the stage: pretraining, adaptation, generation or evaluation.
- Name the inputs, constraints and a case where the approach fails.
Answer
An LLM call is stateless: input → output, done. An AI agent wraps an LLM in a loop:
while not done:
action = LLM(observations, memory, tools)
observation = execute(action)
memory.update(observation)
Key additions:
| Feature | LLM Call | AI Agent |
|---|---|---|
| State | Stateless | Maintains state across steps |
| Execution | Returns text | Executes real actions |
| Iteration | One pass | Multiple reasoning-action cycles |
| Tools | None | Accesses APIs, databases, code |
| Memory | Context only | Short + long-term memory |
Agents unlock tasks impossible with a single LLM call: web research, code execution, database queries, multi-step workflow automation.
Source and licence: JustInCache & contributors — Awesome AI Interviews · Apache-2.0