How to Transfer ChatGPT Memory to Claude (Complete 2026 Guide)

Introduction: Why Moving Your Context Matters

A growing number of users are switching from ChatGPT to Claude for stronger writing quality, better coding and artifact generation, or simply personal preference. The biggest friction has always been context: after months or years of conversations, ChatGPT already knows your work style, projects, preferences, and personal details. Starting over feels painful.

Anthropic solved this with an official Import memory from other AI providers feature. You can now move your accumulated ChatGPT memories into Claude in just a few minutes. The process is simple, works on free and paid plans, and produces clean, categorized memories in Claude that often feel better organized than the original ones.

This guide covers the official method step-by-step, plus practical tips to extract more useful context, how to handle Claude Projects, limitations, and alternatives.

What Actually Transfers (and What Doesn’t)

Claude’s import focuses on memory — the synthesized profile of who you are, your preferences, ongoing work, response style, tools you use, and key context ChatGPT has stored about you.

It does not automatically transfer:

• Full chat history
• Custom GPTs
• Attached files from Projects (in the ChatGPT sense)
• Generated media or deep research reports

Projects need a manual extra step (covered below). Memory imports are still marked experimental by Anthropic, so results can vary, and full integration may take up to 24 hours as Claude runs its nightly synthesis.

Official Method: Transfer ChatGPT Memory to Claude

Step 1: Start the Import in Claude

1. Log in at claude.ai.
2. Click your profile picture (bottom left).
3. Go to Settings → Memory (or Settings → Capabilities on some accounts still on the legacy interface).
4. Find Import memory from other AI providers and click Start import.

Claude shows you a ready-made extraction prompt plus a paste field.
(Alternative direct link that some users report: claude.ai/import-memory.)

Step 2: Prepare ChatGPT for Maximum Extraction

Before pasting the prompt, optimize ChatGPT so it surfaces more context:

1. In ChatGPT, click your profile → Settings → Personalization.
2. Turn on (or confirm they are on):
   • Reference saved memories
   • Reference chat history (or equivalent browser/chat history reference options)
3. Switch the model picker from Auto to a Thinking model. Testing shows the thinking models retrieve more details because they perform extra reasoning cycles over your account data.

If memory features were recently enabled, wait a bit for them to populate.

Step 3: Run the Extraction Prompt in ChatGPT

Copy the prompt Claude provides. It generally looks like this (exact wording can update):

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…)
– Personal details…
– Projects, goals, and recurring topics…
– Tools, languages, and frameworks I use…

Paste it into a new ChatGPT chat and run it. Copy the full output (usually a structured code block).

Step 4: Paste Back into Claude and Import

Return to the Claude import window, paste the entire ChatGPT output into the second field, and click Add to memory.

Claude extracts and formats the information into individual, categorized memory entries. You can review them immediately. Many users note that Claude’s version is cleaner and better organized than ChatGPT’s original memories. Claude also regenerates/updates memories overnight based on new chats.

Optionally, edit any inaccurate entries using the pencil icon.

Handling ChatGPT Projects / Work Context

Global memory is only part of the picture. If you use ChatGPT Projects (or heavy project-style chats with attached files and custom instructions), those live in their own containers.

Recommended approach:

1. Run the same extraction prompt inside the relevant ChatGPT Project (or while referencing that context) using a Thinking model.
2. In Claude, create a matching Project.
3. Add the extracted information to the Project’s memory/instructions.
4. Manually re-upload key files if needed.
5. After Claude’s next memory synthesis cycle (roughly overnight), the Project-level context becomes fully available.

This completes a more thorough migration for power users.

Alternative / Manual Methods

If the official flow isn’t available or you want more control:

• Ask ChatGPT: “Create a data packet summarizing everything you know about me that I can use to give context to another LLM.” Copy the result and paste it into Claude’s preferences or a new chat with instructions such as: “This is my memory from another AI assistant." Add this information into your memory using the memory edit tool.
• Export your full ChatGPT data (Settings → Data controls → Export) for a local archive, then selectively synthesize important parts with Claude or another tool.
• Manually write a concise personal profile covering work, preferences, tools, and current projects and feed that to Claude.

The official import remains the fastest and cleanest for most people.

After the Transfer: Review and Iterate

• Go to Settings → Memory (or View and manage memory) in Claude to see and edit everything.
• Test with a fresh chat: “What do you know about me?” or ask about a specific ongoing project.
• Expect up to 24 hours for full propagation in some cases.
• You can repeat the import later if you accumulate significant new context in ChatGPT.

Claude’s memory is designed to focus more on work-related collaboration context, and it updates continuously from your new conversations.

Limitations and Tips for Best Results

• The feature works with ChatGPT, Gemini, and other memory-enabled AIs.
• Sensitive information: Review the extracted text before importing and remove anything you don’t want transferred.
• Not a perfect 1:1 clone of chat history — think of it as transferring the profile and preferences layer.
• For the absolute richest context, combine the official memory import with selective Project migration and a local data export.

Final Thoughts

Switching from ChatGPT to Claude no longer means losing the personalized context you’ve built up. Anthropic’s built-in import makes the process fast and low-friction. Enable the right ChatGPT settings, use a Thinking model for extraction, handle Projects manually if needed, and you’ll be productive in Claude almost immediately — often with cleaner, better-categorized memories.

Have you completed the migration? Share any extra tips or issues in the comments. If this guide helped, consider sharing it with others making the same switch.

Useful external resources for further reading:

• Official Anthropic help: Import and export your memory from Claude
• Claude memory overview and controls
• OpenAI ChatGPT Personalization / Memory settings documentation

This approach keeps the article actionable, keyword-rich (naturally embedded), structured for scannability and featured snippets, and ready for internal linking to related Claude or AI comparison posts.