What is a multi-turn conversation in AI?
Quick answer
A multi-turn conversation is a chat with several back-and-forth exchanges, where each answer depends on what was said before. "Make it shorter" only works because the AI still has the earlier draft. A single-turn exchange, by contrast, is one question and one answer with no history.
Last updated
Take the free module →Module 6 of the free ChatGPT for Everyday Work course, about 9 minutes. No account needed.
Updated · By Robert Breen
Why it matters for a small business
Language models don't remember anything between requests on their own. A multi-turn chat works because the app sends the earlier messages along with your new one each time. In ChatGPT that happens inside one chat. In an n8n agent, a memory node does it, keeping the last few messages for each chat session.
This explains two everyday surprises. A follow-up like "use 100 guests instead" works in the same chat but not in a new one. And a very long chat can start mixing details, because every old message is still in play. Good multi-turn design means keeping the history that helps and starting fresh when it no longer does.
In a real lesson: Stop Losing Context: When to Start a New Chat vs Keep Going using ChatGPT
In Stop Losing Context, the Harbor Street Bakery office manager pastes Oakwood Library's facts once and gets a quote for 80 guests. When the library calls back, she types one line, "it is 100 guests now, not 80", and ChatGPT redoes the quote. Then she asks for a text to Sam in the same chat, and it uses the new headcount.
That is three turns building on each other. For the next customer, Pine Ridge Rotary, she clicks New chat and attaches only their facts. In Book Appointments from Chat, the agent's memory node does the same job, so a customer can give their name in one message and the time in the next.
Try this lesson free or read the step-by-step guide.
Common confusions
Multi-turn vs memory
Multi-turn means the earlier messages of this chat are sent along. Memory features save notes across separate chats, which is a different system.
Multi-turn vs multi-agent
Multi-turn is one AI and one person trading messages. Multi-agent means several AI agents working together on a task.
Tips
- Fix a draft with short follow-ups in the same chat instead of rewriting the whole prompt.
- Start a new chat for a new customer or task, so old details can't leak in.
- In an n8n agent, add a memory node, or each message is answered as if it were the first.
Related terms
More AI assistants terms
Where you use it: free lessons
- Stop Losing Context: When to Start a New Chat vs Keep Going using ChatGPT (ChatGPT, 9 min)
- n8n AI Agent with Tools: Check Google Calendar, Book It and Log It (n8n, 11 min)
- Chatbase: Answer Customer FAQs with an AI Agent (Chatbase, 15 min)
Frequently asked questions
- Does the AI remember earlier messages in a chat?
- Yes, within one chat, because the app sends the earlier messages with each new one, up to the model's context window.
- Why did my n8n agent forget what I said a moment ago?
- It probably has no memory node, or the session ID changed. Without memory, each message reaches the model on its own.
- What is a turn?
- One message from one side. A person's question and the AI's reply are two turns.
- Do longer chats cost more?
- Usually yes when you pay per token, because the earlier messages are sent again with every new turn.