Planner-Executor

Full definition

Unlike ReAct (which interleaves thought and action one step at a time), planner-executor commits to a plan, then executes it, then optionally re-plans on failure. Variants include LLM-Compiler (parallel execution of independent steps) and ADaPT (adaptive decomposition). Planner-executor is preferred for long-horizon tasks (>10 steps) because the planner can use a slower, smarter model while the executor uses a faster, cheaper one — splitting cost from quality.

Why it matters

For multi-step enterprise workflows (research, data ETL, code refactoring), planner-executor delivers higher reliability than ReAct at lower total cost — the planner runs once, the executor many times. Procurement teams should ask agent vendors which pattern they use and why.

Example

A market-research agent uses GPT-4 as planner ("1. find competitors, 2. extract pricing, 3. summarize, 4. format as table") and Llama 3 8B as executor for each step; total cost drops 70% vs all-GPT-4.

Related terms

• AI Agent / Agentic AI
  An AI agent is an LLM-driven system that, given a goal, plans a sequence of steps, calls tools or APIs, observes the results, and iterates until the goal is reached — going beyond single-turn chat to multi-step autonomous action.
• ReAct Pattern
  ReAct (Reasoning + Acting) is a prompting pattern that interleaves reasoning steps ("Thought") with tool actions ("Action") and tool results ("Observation"), letting an LLM iteratively decompose a task, query the world, and self-correct.
• Agentic Workflow
  An agentic workflow is a multi-step AI process where a model plans, takes actions through tools, observes results, and iterates toward a goal — replacing single-shot prompting with a loop that reasons, acts, and self-corrects.
• Tool Router
  A tool router is a layer in an agent system that selects which tool (or subset of tools) to expose to the model for a given query, instead of putting all tool definitions in every prompt.