ChatGPT guide

How to continue a ChatGPT conversation after hitting the limit

When you hit ChatGPT's context window, you get either a hard error or silent forgetting. Here's how to continue without losing your work — and how to make sure it never disrupts you again.

Start free on Chrome

Why you can't just keep going

  • ChatGPT doesn't have unlimited memory. Every conversation has a context window — a maximum amount of text the model can hold in working memory at once. When the limit is reached, either ChatGPT shows a hard error or silently drops the oldest messages, causing it to forget earlier context.
  • The silent version is more frustrating because you don't realise it's happening until responses start ignoring things you already established.

How to continue after hitting the limit

  • Ask ChatGPT to summarise everything covered so far, including key decisions, outputs, and what still needs doing.
  • Copy that summary and open a new ChatGPT conversation.
  • Paste the summary as your first message and pick up from there.
  • For complex sessions with code or documents, also paste the final version of any outputs directly into the new chat.

The one-click fix: thredly

  • Every manual method puts the burden on you. thredly is a Chrome extension that adds a Summarize button inside ChatGPT, Claude, and DeepSeek. Open the long thread, click Summarize, paste the structured handover into a fresh chat. One click instead of writing the summary yourself. It only runs when you click, and summaries are private to your account.

Related guides

Frequently asked questions

Does starting a new conversation delete my old one?

No, your old conversation is still accessible in your sidebar. You're just beginning a new session. ChatGPT won't carry context across on its own, so you need to bring it yourself, or use thredly's one-click Summarize button to generate the handover.

Will ChatGPT warn me before I hit the limit?

Usually not before it's too late. You may notice response quality degrading before the hard error. This is why proactive summarisation matters.