In partnership with
There is one reason most people stay on ChatGPT longer than they should.
It is not the features. It is not the price. It is context debt.
You have spent months, maybe years, teaching a model who you are. Your tone, your projects, your quirks, your preferences. Starting over somewhere else feels like throwing all of that away. And that sunk cost is real.
But it does not have to be. The full migration memories, conversation history, and Custom GPTs takes less than 25 minutes. Here is exactly how to do it.
Before You Start
Save Your Custom Instructions First
Before running any export prompt, go to ChatGPT Settings > Personalization and manually copy your Custom Instructions into a separate document.
These are the standing rules you wrote: tone preferences, output formats, things it should always or never do. They live separately from stored memories and will not be captured by the export steps below.
Once copied, paste them into Claude at Settings > General > Personal Preferences. That is your foundation. Now move on to the memory transfer.
Step 1: The 2-Minute Memory Import
Open ChatGPT and paste this prompt exactly as written:
Export all of my stored memories and any context you've learned about me from past conversations. Preserve my words verbatim where possible, especially for instructions and preferences.
Categories (output in this order):
Instructions - Rules I've explicitly asked you to follow going forward: tone, format, style, "always do X," "never do Y," and corrections to your behavior. Only include rules from stored memories, not from conversations.
Identity - Name, age, location, education, family, relationships, languages, and personal interests.
Career - Current and past roles, companies, and general skill areas.
Projects - Projects I meaningfully built or committed to. Ideally ONE entry per project. Include what it does, current status, and any key decisions. Use the project name or a short descriptor as the first words of the entry.
Preferences - Opinions, tastes, and working-style preferences that apply broadly.
Format: Use section headers for each category. Within each category, list one entry per line, sorted by oldest date first. Format each line as: [YYYY-MM-DD] - Entry content here. If no date is known, use [unknown] instead.
Output: Wrap the entire export in a single code block for easy copying. After the code block, state whether this is the complete set or if more remain.
Copy the full code block it returns.
Then go to Claude: Settings > Capabilities > Memory > Start Import. Drop your export into the text box and click "Add to Memory."
Claude reads the block, extracts the key information, and stores it as individual rules. Your new assistant now knows what your old one knew. That's it. The sunk cost argument is gone.
AI Alone Can’t Run Revenue
Finance doesn’t run on “mostly right.” It runs on math.
In The Architecture Behind AI-Native Revenue Automation, Tabs’s CTO breaks down why LLMs alone aren’t enough—and what it actually takes to build audit-ready, AI-driven contract-to-cash systems for modern B2B teams.
Step 2: The 20-Minute Deep Export With Cowork
The memory import is fast but shallow. It captures what ChatGPT stored, not what actually happened across your conversations. If you have been on the platform for two or more years, the real context is buried inside your conversation history.
Go to ChatGPT Settings > Data Controls > Export Data. You get a zip file containing conversations.json. Every message you have ever sent is in that file. One person's export was 267MB covering nearly 3,000 conversations across 18 months.
Drop that file into Cowork and run this prompt:
Read conversations.json (full ChatGPT export). Process every conversation. Produce a single markdown file called profile.md.
This is NOT a list of memories. This is a comprehensive profile document. Write it as a reference brief that another AI assistant would read to fully understand who I am, how I work, and how to write for me. Use complete sentences. Be specific. Include exact numbers, names, dates, and frameworks wherever they appeared in conversations.
Then ask it to structure the document with these sections:
WHO I AM / WHAT I DO / WHO I WORK WITH / HOW I WRITE / MY FRAMEWORKS AND SYSTEMS / MY CONTENT STRATEGY / MY TOOLS AND TECH STACK / ACTIVE PROJECTS AND GOALS / KEY NUMBERS / RECURRING PROBLEMS / OPINIONS AND PREFERENCES / TIMELINE
A few rules to include in the prompt: write in third person, no filler, deduplicate repeated facts, target 5,000 to 15,000 words, and output as a single markdown code block.
Cowork reads the file, runs a script to parse every conversation, and returns a structured document. What comes back is years of context compressed into a single reference you can use anywhere.
1,000+ Proven ChatGPT Prompts That Help You Work 10X Faster
ChatGPT is insanely powerful.
But most people waste 90% of its potential by using it like Google.
These 1,000+ proven ChatGPT prompts fix that and help you work 10X faster.
Sign up for Superhuman AI and get:
1,000+ ready-to-use prompts to solve problems in minutes instead of hours—tested & used by 1M+ professionals
Superhuman AI newsletter (3 min daily) so you keep learning new AI tools & tutorials to stay ahead in your career—the prompts are just the beginning
Step 3
Three Ways to Use Your profile.md
Import it at Settings > Capabilities > Memory. Claude references it automatically across every conversation you open.
Save it to your Cowork folder. Claude reads it at the start of every Cowork session alongside your other .md files, giving it the depth it needs for longer tasks.
Upload it to a Claude Project. Every conversation inside that project starts with your full context already loaded. No need to re-explain yourself.
You do not have to pick one. Using all three covers different layers: Memory for the basics everywhere, Cowork for deep work, Projects for focused tasks.
Step 4: Convert Your Custom GPTs Into Claude Skills
The memory and profile handle your context. Custom GPTs handle your workflows. Bring those across too.
Open any Custom GPT you have built. Go to Configure and copy the full system instructions. Download any files you uploaded to it. Then paste this into Claude:
I'm going to give you the instructions from a Custom GPT I've built. Reformat this as a Claude Skill. Keep all the logic, rules, and behaviour intact. Then list every file I should attach so the skill has the same context the GPT had.
Paste the instructions below the prompt. Claude rebuilds it as a reusable skill. Same logic, same rules. You are not starting over. You are just moving it.
What To Do With All That Context
Once you have everything transferred, create a Claude Project and call it "My Second Brain." Upload your profile.md, your context files, and any documents you reference regularly. This becomes your personal knowledge base.
Then ask it something specific. Ask it to pull everything you ever said about how you grew an audience, built a product, or handled a recurring problem. Structure it as a framework. What comes back will surprise you. You will find patterns in your own thinking you had forgotten were there.
Close the Loop
From Context to Content to Branded Visual
Once your old context lives inside a Claude Project, you can turn it into content fast. Pull a framework from your conversation history, shape it into a post, then drop it into Vislo AI to generate a branded LinkedIn infographic in seconds.
Vislo keeps your fonts, your colours, and your layout intact. No regenerating because a logo shifted or text broke. Everything stays editable.
Vislo is now in public beta with the first 100 spots open. If you want early access, sign up here. There is also a walkthrough video showing the full workflow.
See Why HubSpot Chose Mintlify for Docs
HubSpot switched to Mintlify and saw 3x faster builds with 50% fewer eng resources. Beautiful, AI-native documentation that scales with your product — no custom infrastructure required.
Quick Reference
The Complete Migration Checklist
- Copy Custom Instructions from ChatGPT and paste into Claude Personal Preferences
- Run the memory export prompt and import the code block into Claude Memory
- Request your full data export from ChatGPT (Settings > Data Controls)
- Drop conversations.json into Cowork and generate your profile.md
- Use profile.md in Claude Memory, Cowork, and a Claude Project
- Copy your Custom GPT instructions and convert each one to a Claude Skill
- Build your "My Second Brain" Project and start pulling structured frameworks from your history
Every step in this guide is tested. Every prompt is copy-paste ready.
The only thing left to do is run it.
Thanks for reading. If this was useful, share it with someone still sitting on the fence about switching. And if you have been here a while, your continued readership is what keeps this newsletter going.
