Copy the extraction prompt into ChatGPT

Open a new chat in ChatGPT and paste this prompt. It's based directly on Anthropic's own official prompt for pulling memory out of ChatGPT — the same one Claude's import tool gives you.

I'm moving to another service and need to export my data. List every memory you have stored about me, as well as any context you've learned about me from past conversations. Output everything in a single code block so I can easily copy it. Format each entry as: [date saved, if available] - memory content. Make sure to cover all of the following - preserve my words verbatim where possible: Instructions I've given you about how to respond (tone, format, style, 'always do X', 'never do Y'). Personal details: name, location, job, family, interests. Projects, goals, and recurring topics. Tools, languages, and frameworks I use. Preferences and corrections I've made to your behavior. Any other stored context not covered above. Do not summarize, group, or omit any entries. If any of this reflects a background summary of our past conversations rather than something explicitly stated, describe your current best understanding of it rather than forcing an exact quote. If you have no memories or stored context about me at all, respond with exactly: "No memories found." After the code block, confirm whether that is the complete set or if any remain.

Ask ChatGPT to clean up its own reply

As your next message in the same conversation, paste this. It asks ChatGPT to reformat what it just told you — fixing awkward third-person phrasing, flagging anything that might be outdated, organizing it into clear sections — entirely inside your own chat. Nothing here touches our servers.

Thanks. Now please reformat what you just listed so I can hand it to another AI assistant. Please:

- Rewrite anything phrased about me in the third person (e.g. "The user prefers...", "User is a...") into direct statements about me (e.g. "Prefers concise answers", "Works as a software engineer").
- Organize everything under these section headers, in this order, only including headers that actually have content under them:
  ## Instructions
  ## Identity
  ## Projects & Goals
  ## Tools & Skills
  ## Preferences
  ## Other
- If anything looks time-sensitive or possibly outdated (an upcoming trip, an unfinished project, a stated age or life stage), keep it but add a short inline note like "(may be outdated — verify)" rather than stating it as current fact.
- Merge duplicate or clearly overlapping entries into one, but don't drop a distinct fact just to shorten the list.
- Don't add, infer, or invent anything that wasn't in your previous reply — this is a reformatting pass only.
- Output the result as a single Markdown code block so I can copy it in one go, with no extra commentary before or after it.

If your previous reply was empty or said there were no memories, just reply with exactly: "No stored memory to format." and stop there.

Copy the final result

Copy ChatGPT's reformatted reply. That's your portable memory profile — ready to paste into the import guide for whichever service you're moving to.

Open Gemini's import page

Google built a page specifically for bringing your context in from another AI service.

Paste your profile, or upload your chat export

Paste your already-formatted memory profile directly, or — uniquely among the services we cover — upload a full .zip chat-history export instead. Gemini will process either.