AiClubb
Expert0 prompts

ReAct Prompting

ReAct (Reasoning + Acting) prompting interleaves reasoning steps with actions. The model reasons about what it needs, acts to gather information (search, compute, call a tool), observes the result, and repeats until it can answer. It powers modern agentic AI systems.

How It Works

1

Thought: the model states what it needs to know.

2

Action: it performs a tool call (search, API, calculator).

3

Observation: it reads the tool result.

4

Repeat Thought → Action → Observation until a final answer is produced.

Real Example

Question: "What is the population of the capital of France?" Thought: I need the capital first. Action: search("capital of France"). Observation: Paris. Thought: now find population. Action: search("Paris population"). Observation: ~2.1M. Final Answer: 2.1 million.

Best Practices

  • Make tools available and named clearly (search, calculator, database).
  • Ask the model to think before every action.
  • Instruct it to stop when enough evidence is gathered.
  • Handle tool failures gracefully: retry or reformulate.

Common Mistakes

  • Actions without reasoning produce wasted tool calls.
  • Missing observations between actions.
  • No stopping condition causes infinite loops.

When to Use ReAct

Agentic workflowsResearch assistantsSystems with tool access

Prompts using ReAct Prompting

Ready-to-copy examples of this technique in action.

No prompts found. Try adjusting your filters.

Frequently Asked Questions

When should I use ReAct Prompting?

Agentic workflows, Research assistants, Systems with tool access. If your task is more complex or your output is inconsistent, consider combining it with examples or chain-of-thought.

Does ReAct Prompting work on all AI models?

Most prompting techniques transfer across modern models like ChatGPT, Claude, and Gemini, though results vary. Test the same prompt on two models and keep the best performer.

What is the next technique to learn after ReAct?

Based on your related techniques, explore: Chain of Thought, Tree of Thoughts, Agent Prompting.

How do I practice ReAct?

Use the prompts above as starting points, then modify one variable at a time. Track which changes improve your results in a prompt library.

Skip to content