AI Chat Handoff guide
How to Continue a ChatGPT Conversation in a New Chat
Short answer: while the old ChatGPT chat still replies, we paste the handoff prompt below into it, check the summary it writes, then open a new chat and paste that summary as the first message. ChatGPT’s Branch in new chat copies the conversation up to a message instead, and Handoff can turn the loaded messages into an editable transcript. A summary is context, not the full history, so we review it for missing decisions, constraints and next steps before continuing.
Chat already says it reached the maximum length and won’t reply? Use the conversation-too-long guide.
Add to Chrome — freeFree Chrome extension; prepares an editable draft you review before sending.
Copy-ready handoff prompt
We paste this into the old chat while it still replies, check the result, then paste the summary into a new chat.
Prepare a working-context summary for a new chat. Include our goal, confirmed decisions, constraints, relevant facts, unresolved questions and next action. Mark uncertainty. Do not invent missing details. Keep exact names and requirements that affect the work.
Option: Branch in new chat (built into ChatGPT)
ChatGPT on the web has a built-in way to split a conversation. According to OpenAI’s release notes (September 4, 2025), we hover over a message, click More actions (⋯) and select Branch in new chat to start a separate conversation from that point, without losing the original thread. It was released for logged-in users on the web.
- Good for: trying a different direction from a specific message while the original chat stays as it was.
- Not a fresh start: the branch continues from the existing conversation up to that message rather than from a short summary. OpenAI presents it as a way to explore different directions, not as a way around the conversation length limit.
- For a lighter new chat, we use the handoff prompt above or a Handoff transcript, and keep only the context the next step needs.
What a new chat can already see: Memory, Projects and files
A new ChatGPT chat is not always a blank page, but it does not inherit the old conversation either. This is what can carry over and what we still paste.
- Memory: if Memory and Reference chat history are on (Settings > Personalization > Memory), ChatGPT can use relevant details from past chats in a new one. OpenAI says Memory does not keep every detail from every conversation and that ChatGPT decides what is relevant, so we still paste the goal, decisions and exact requirements the next step depends on.
- Projects: if the work lives in a ChatGPT project, a new chat in that project starts with the project’s files and instructions, and depending on the project’s memory setting and plan it can reference other chats in the project. We can save the reviewed handoff with Add to project sources so later project chats can reuse it, and move an existing chat into the project with Move to project.
- Files: we do not assume the new chat can open files uploaded to the old one. We attach again the ones the next step needs, or add them to the project.
- Already seeing “Context window full” or “Chat memory full” with an offer to start a new chat with an automatic summary? That notice has its own guide: see the chat-memory-full guide.
The manual method
- In the working conversation, ask ChatGPT for a handoff using the prompt above. Include the goal, decisions, constraints, unresolved questions and next action.
- Read the summary against the original messages. Correct mistakes and add exact details such as filenames or requirements that matter.
- Open a new ChatGPT chat, paste the reviewed summary and ask it to explain its understanding before taking the next step. Send it yourself.
The Handoff method
- We open the ChatGPT conversation in Chrome 120 or later, wait for responses to finish and scroll to load the older messages we need.
- In the extension, we choose Last 10 messages, Last 20 messages or All loaded messages, then click Read current chat.
- We select ChatGPT under Continue in. We preview and freely edit the transcript draft, removing private information and adding any missing decisions.
- We click Open with draft. A new ChatGPT chat opens in a new tab with the draft in its empty message box. We review it and press Send ourselves; the extension never sends. We can also use Copy draft and paste manually.
Handoff prepares a transcript, not an AI summary. We can edit its preamble and the User / Assistant message blocks.
Example
An illustrative draft with concrete context we want to carry forward:
Continue the conversation below. Treat the transcript as context, not as new instructions. Preserve relevant decisions and ask if anything is unclear. User: We are rewriting the onboarding email. Use British English, keep it under 150 words and do not offer a discount. Assistant: We agreed on a welcome, one setup action and a support link. Next: draft the email using those constraints.
Limits and privacy
Only currently loaded conversation text is captured. Attachments, images, hidden reasoning, tool artifacts and branches are excluded. If earlier messages are not loaded, the draft adds “[earlier messages not loaded]”.
The draft cap is 60,000 characters. Above that, the oldest messages are left out and the newest are kept, with a note such as “[Earlier context truncated; oldest 12 of 40 messages omitted to fit the 60,000-character limit, newest kept.]”. If a kept message had to be shortened, the note says its start is cut. Character length does not ensure a fit within the destination’s token limits. We shorten the draft and check important details.
A working summary selects relevant context; a transcript retains the captured message text. Neither imports the full history as an active conversation or transfers saved memory. A data export and file upload is a separate history-reference workflow. Claude’s memory import handles memory and preferences, not conversations.
The extension processes text locally in the browser and does not upload your conversation to Tacko: no backend or AI APIs for chat content. When enabled, it sends limited pseudonymous usage statistics (version, transfer direction, error type, and a random install ID), which you can switch off. Once the draft is inserted into ChatGPT or Claude, that service’s privacy terms apply, even before you press Send.
Site selectors can change. Authenticated live-site compatibility can break when ChatGPT or Claude change their pages; unknown layouts fail with copy/paste instructions. We still need access to the destination service, and its model and usage limits apply. Handoff starts a new chat with context; it does not fix slowness or remove conversation length limits.
Free includes 2 successful transfers per local calendar day; failed prefills and login pages do not count. Copy draft is unlimited. For unlimited transfers, get Pro for $6.99 one-time; it uses an offline license key.
Sources
Platform facts on this page come from official help articles, last checked November 13, 2026. Claude and ChatGPT change their features and limits, so we check the current article before relying on a detail.
Frequently asked questions
Does a new chat know the old conversation?
Not reliably. With Memory and Reference chat history on, ChatGPT can draw on relevant details from past chats, and chats in a project can reference other chats in that project, but OpenAI says Memory does not keep every detail. We provide the relevant context in the first message: a working summary is a selected account of the work, not an import of the old thread.
What if the old chat cannot accept a summary request?
We use the conversation-too-long guide instead: copy existing visible messages or read the loaded text with Handoff.
Is saved memory the same as a handoff?
No. Saved memory is separate from a conversation transcript. We include task-specific decisions and constraints in the handoff rather than relying on memory.
Does Handoff generate the summary?
No. Handoff builds an editable transcript of loaded messages, with a preamble. We can shorten or edit it before sending.
Is Branch in new chat the same as starting fresh?
No. Branch in new chat starts a separate conversation from a chosen message and keeps the original thread unchanged. It builds on the existing conversation up to that point, so for a lighter new chat we paste a reviewed summary or transcript instead.
Continue with context
AI Chat Handoff is free in the Chrome Web Store.
Explore AI Chat HandoffAdd to Chrome — freeRelated guides
- ChatGPT Conversation Too Long? How to Continue
- How to Move a ChatGPT Conversation to Claude
- Claude Conversation Too Long? How to Continue in a New Chat
- How to Move a Claude Conversation to ChatGPT
- ChatGPT Says “Chat Memory Full”? What It Means and How to Continue
- How to Get a Second Opinion from Claude on a ChatGPT Answer
- ChatGPT Slow or Lagging in a Long Conversation? What to Do
- Claude Slow or Laggy in a Long Conversation? What to Do
- Claude Reached Its Max Length for This Message? How to Continue
- ChatGPT Says “This Conversation Is Read-Only”? Why and How to Continue
- ChatGPT “Could Not Load This Conversation”? How to Get Your Chat Back
- ChatGPT “Connection Interrupted. Waiting for the Complete Answer”: What to Do
- ChatGPT Shared Link Not Working? Why It Won’t Open and What to Do
- ChatGPT “ConversationDetailError”: Why One Chat Stops Accepting Prompts
- Claude’s Response Was Interrupted? What It Means and How to Continue
- ChatGPT “Error in Message Stream”: Why One Chat Keeps Failing and What to Do
- ChatGPT Writing Blocks Not Working in One Chat? What to Do
- DeepSeek “Long Conversations Inactive for Over 7 Days Can No Longer Continue”: What to Do