Not "important" like saving the world. Important like — I was finally in a flow. Claude was helping me think, write, plan. And then it just... stopped.

"You've reached your usage limit."

I stared at the screen. I had two options. Pay extra. Or figure out what the hell was actually happening.

I chose the second one. Not because I'm cheap — okay, maybe a little — but because it felt wrong to pay more without understanding why I was running out in the first place.

So I did what I do. I went looking.

I read Anthropic's official help docs. I watched reels. I commented on threads and actually got some really useful answers back. I posted my own question in my community. I went down a GitHub rabbit hole at 11pm when I should have been sleeping.

And what I found genuinely surprised me.

It wasn't my prompts. It wasn't that I was asking too much. The limit was running out because of things I didn't even know were happening in the background — silently, every single message, before I even typed a word.

This blog is everything I found. All in one place. Because I don't want you to have to do what I did.

🤯 What's Actually Happening Behind the Scenes?

Most people think they hit the limit because they asked too many questions or wrote too much. That's not really it.

Here's the thing nobody tells you: every single message you send, Claude silently reloads your full conversation history, your memory profile, your connected tools, and your style settings in the background. Thousands of tokens — gone before you even typed your question.

A creator named _prem.io posted a viral thread breaking down exactly what was eating his tokens in a 20-message chat:

Every. Single. Chat.

After fixing his settings, he went from ~198,000 tokens to ~10,000 tokens per chat. That's a 94% reduction — same output, same quality.

💡 The limit isn't running out because you're doing too much. It's running out because the settings were silently working against you.

📖 Official Sources I Read (So You Can Too)

• 📄 Anthropic Help: How do usage and length limits work? 👉 https://support.claude.com/en/articles/11647753-how-do-usage-and-length-limits-work

• 📄 Anthropic Help: Usage limit best practices 👉 https://support.claude.ai (search "Usage limit best practices")

• 📄 Anthropic Help: What is Claude's memory? 👉 https://support.claude.ai (search "Claude memory")

• 📄 Anthropic Help: Searching past chats with Claude 👉 https://support.claude.ai (search "searching past chats")

• 🐙 GitHub: Caveman Claude Skill (cuts output tokens ~75%) 👉 https://github.com/amanattar/caveman-claude-skill

• 🐙 GitHub: Caveman CLI Tool (#1 trending on GitHub, 5K+ stars) 👉 https://github.com/JuliusBrussee/caveman

• 🐙 GitHub: Code Review Graph (for developers, 12K+ stars) 👉 https://github.com/tirth8205/code-review-graph

⚙️ Part 1: Settings to Change Right Now (Do This Once)

1. Turn Memory OFF

Settings → Memory → toggle off

Every message loads your full memory profile silently. Turning this off saves ~60,000 tokens per 20-message chat instantly.

Still need context? Use Notion. Create 3 lean pages (your identity, platform rules, active project). Start each chat with: "Load my profile from Notion." Claude fetches it once — not every message.

💰 Saves ~60,000 tokens per 20-message chat

2. Switch Style to "Concise"

Tap + icon before sending → Use style → Concise

Claude's default style is verbose. Switching to Concise cuts output tokens by roughly half — automatically, no prompting needed.

💰 Cuts output tokens by ~50%

3. Use Sonnet, Not Opus, for Everyday Work

Opus is powerful but expensive on your token budget. For captions, content drafts, brainstorming — Sonnet does the same job. Save Opus for when you genuinely need deep reasoning.

💰 Sonnet is up to 5× more token-efficient than Opus

4. Turn Web Search OFF When You Don't Need Live Data

Tap + icon → toggle web search off

Web search is ON by default. For writing, planning, or creative tasks — it's adding overhead to every single message for no reason.

5. Set Connected Tools to "On Demand"

Each connected tool (Notion, Google Drive, Slack) loads its full schema into every message — even when you're not using it. That's 2,000–5,000 extra tokens per message.

Fix: Go to tool settings → change from "Always loaded" to "On demand"

💬 Part 2: Conversation Habits to Change

6. Load Your Context in the First Message

Anthropic officially recommends this. Before you start — plan what you need, what background you're providing, what you're asking. Put it all in the first message. Reduces back-and-forth significantly.

7. Edit Your Prompt — Don't Send a New One

Made a mistake? Click the edit icon on your last message instead of sending a follow-up. A new message = Claude reloads the full conversation. Editing = much cheaper. Almost nobody knows this one.

8. Batch Your Questions — One Message, Not Three

Instead of three separate messages, write it all in one:

"Summarise this article, list the key points, and suggest a headline."

Anthropic confirms this in their usage guide. Every separate message reloads your full context.

9. Start a Fresh Chat After ~10 Messages

Long conversations drag full history into every message. After ~10 exchanges, open a new chat. Use Projects to carry forward only what you actually need.

10. Put Your Rules in a Project — Once

Anthropic confirmed: content in Projects is cached and doesn't count against limits the same way when reused.

Add your instructions, constraints, and background context once. Claude follows them every chat without you repeating yourself.

💰 Cached project content = reused without full token cost

11. Don't Upload Heavy Files — Copy-Paste Instead

A single PDF page costs 1,500–3,000 tokens per Anthropic's own docs. A full screenshot costs 1,300+ tokens. Extract just the text you need and paste it plain. Crop screenshots tight.

12. Use Claude During Off-Peak Hours

Anthropic officially confirmed: during weekday peak hours (5am–11am Pacific / ~6:30pm–12:30am IST), your session limit burns faster. Evenings and weekends give you significantly more room. Schedule heavy Claude work accordingly.

🐙 Part 3: GitHub Tools Worth Knowing

🪨 Caveman Mode — Make Claude Talk Less

👉 https://github.com/JuliusBrussee/caveman (5K+ stars, was #1 trending on GitHub) 👉 https://github.com/amanattar/caveman-claude-skill

Make Claude respond in compressed, caveman-style language. No filler. No pleasantries. Just the answer. Technical substance stays completely intact.

Normal Claude:

"The reason your React component is re-rendering is likely because you're creating a new object reference on each render cycle..."

Caveman Claude:

"New object ref each render. Inline object prop = new ref = re-render. Wrap in useMemo."

Same information. 75% fewer tokens. There are intensity levels — lite, full, ultra, and even classical Chinese compression modes if you're feeling adventurous 😂

The SKILL.md file can be added to any Claude Project as an instruction — so this isn't just for developers.

📊 Code Review Graph — For Developers

👉 https://github.com/tirth8205/code-review-graph (12,000+ stars)

Claude re-reads your entire codebase on every task. This tool builds a persistent map so Claude only reads what's relevant. Results: up to 8x fewer tokens on code reviews, up to 49x on daily coding tasks. Not for beginners — but worth knowing it exists.

🧠 My Honest Take

Most people blame their prompts when they hit the limit.

The real culprit is the invisible overhead — memory loading, tool schemas, web search running silently in the background, verbose responses adding up.

Fix the settings first. Change your conversation habits second. Prompts are the last thing you need to worry about.

I went from hitting limits mid-week to finishing the week with usage left — just by changing settings. No better prompts. No upgrade. Just the right switches flipped.

✅ Quick Checklist — Save This

Settings (do once):

• Memory OFF → Settings → Memory

• Style → Concise (+ icon)

• Sonnet for everyday tasks

• Web search OFF when not needed

• Tools → "On Demand"

Habits:

• Heavy context in your first message

• Edit prompts, don't send new ones

• Batch questions in one message

• New chat after ~10 messages

• Rules inside a Project — once

For developers:

• Caveman Mode → https://github.com/amanattar/caveman-claude-skill

• Code Review Graph → https://github.com/tirth8205/code-review-graph

📎 All Links

Written by Ranjani Shetty — AI Educator helping content creators, students, and brand owners learn AI practically. Follow me on Instagram: ranjani shetty

If this helped you, share it with one person who's also hitting the Claude limit. They'll thank you. 🙏

For almost a month, I was running a comment strategy on Instagram — someone would comment on my reel, and I'd DM them a Google Doc link with all the information.

It worked. But it felt like jugaad.

A random Google Doc link in someone's DMs doesn't say "professional creator." It says "I figured this out five minutes ago."

So I fixed it. I built my own website. In under 10 minutes. For ₹99.

Here's exactly how I did it — step by step, so you can do it too.

What You Need (Before We Start)

Three tools. All free except the domain:

• GoDaddy — to buy your domain (₹99)
• Hashnode — to write and publish your blog (free)
• Superprofile — your link-in-bio that looks like a real website (free)

Step 1 — Buy Your Domain on GoDaddy

Go to godaddy.com and search for your name or brand name.

Tips for picking a good domain:

• Keep it short and easy to spell — no one should have to guess it
• Use your real name if possible (mine is ranjanishetty.in)
• Pick .in for an Indian audience, .com if you want to go global
• Avoid numbers, hyphens, and weird spellings
• If your name is taken, try adding your niche — like ranjaniai.in

Once you find a good one — add to cart and check out. I paid ₹99 for mine.

You now own your corner of the internet. That feels good, trust me.

Step 2 — Set Up Your Free Blog on Hashnode

Hashnode is where you write your blogs — tutorials, prompts, guides, tips. It looks clean, professional, and Google can find it. And it's completely free.

Go to hashnode.com and create a free account.

Setting up your blog:

1. Add your profile photo and a short bio
2. Give your blog a name (e.g. Ranjani Shetty | AI for Everyone)
3. Write your first post — even a short intro post works
4. Hit publish — your blog is now live on hashnode.dev/yourusername

Connecting your GoDaddy domain to Hashnode:

1. Go to your Hashnode dashboard → Settings → Domain
2. Type in your domain (e.g. ranjanishetty.in) and save
3. Go to GoDaddy → My Products → DNS Settings for your domain
4. Add a CNAME record: Name = www, Value = hashnode.network
5. Save it. Wait 10–30 minutes for it to connect.
6. Done — your blog now lives at your own domain!

It sounds technical but it's really just copy-paste. If I did it as a toddler mom with 3 hours a day, you can definitely do it.

Step 3 — Set Up Your Superprofile

Think of Superprofile as your link-in-bio — but one that looks like an actual website. Instead of sending people to a plain Instagram bio link, you send them somewhere that shows your blog, your offers, your WhatsApp, everything — in one place.

Go to superprofile.bio and create a free account.

What to add to your Superprofile:

• Your name, photo, and one-line bio
• Link to your Hashnode blog
• Your Instagram handle
• WhatsApp link for direct enquiries
• Any offers — free guides, prompt packs, cohort waitlist
• About section — who you are and who you help

Once it looks good — copy your Superprofile link and paste it in your Instagram bio.

Now your Instagram bio has one link. That one link takes people to your whole world.

Step 4 — Connect Everything

Here's how the whole thing flows:

• GoDaddy domain → points to your Hashnode blog
• Hashnode blog → linked inside Superprofile
• Superprofile → sits in your Instagram bio
• Instagram → sends followers to Superprofile → they find your blog, offers, WhatsApp

One link. Everything in one place. Looks like a real brand. Because it is.

What to Add to Make Your Website Even Better

Now that your website is live, here's what most beginners miss:

• A proper About page — Tell your story. Who you are, what you teach, why you started. People buy from people they trust.
• A free resource or lead magnet — A free prompt pack, checklist, or mini-guide. Give something away so people have a reason to come back.
• A contact or WhatsApp link — Make it easy for people to reach you. If they have to search, they won't bother.
• Consistent blog titles with keywords — Write titles like "How to use Claude for Instagram" not "My AI Journey." This is how Google finds you.
• A newsletter signup — Hashnode has a built-in newsletter. Turn it on. Even 50 subscribers is an asset.
• Screenshots and visuals inside blogs — Walls of text are boring. Add screen recordings and screenshots — make it visual.
• Internal links between blogs — Link one blog to another. Keeps readers on your site longer and helps SEO.
• A consistent posting schedule — One blog a week is enough. Consistency beats frequency every time.

You Don't Need a Developer. You Need 10 Minutes.

Let's recap:

• GoDaddy domain — ₹99
• Hashnode blog — ₹0
• Superprofile — ₹0
• Total — ₹99 and 10 minutes

I did this as a toddler mom with 3 hours a day. If I can do it — so can you.

Want more practical AI and creator tips like this?

Follow me on Instagram → @ranjanicshetty Read more blogs → ranjanishetty.in

Ranjani Shetty is an AI educator with 10 years of ecommerce experience. She teaches creators, students, and working moms how to use AI tools practically — one reel at a time.

I didn't hire a social media manager. I built one. In one prompt.

Here's exactly what I did, what Claude gave me back, and the full prompt you can steal today — whether you're a fitness coach, CA student, small business owner, or anyone trying to grow online.

Why Most People Get Bad Output From AI

They type something like:

"Help me plan my Instagram content"

And they get something generic, useless, and forgettable.

The problem isn't Claude. The problem is the input.

Claude is only as good as what you give it. Vague in = vague out. Specific in = specific manager-level output.

Before you paste any prompt, you need to answer 10 questions about yourself. Think of this as your briefing document — the same thing you'd hand a real social media manager on day one.

Answer These 10 Questions First

Work through these before you open Claude. The clearer your answers, the sharper your output.

Q1. What is your niche and what do you teach or offer?

Bad answer: "I make content about AI"

Good answer: "I teach working moms and students how to use Claude and ChatGPT for productivity and income — in simple, jargon-free Hinglish"

Q2. Who exactly is your audience?

Bad answer: "Everyone interested in AI"

Good answer: "Indian women aged 22–35, working or studying, basic smartphone users, not techies, want practical AI skills to save time or earn money"

Q3. What is your content tone?

Bad answer: "Friendly"

Good answer: "Like a didi explaining something over chai — warm, simple, zero jargon, occasionally funny, always actionable"

Q4. What are your 3–4 content pillars?

These are the core themes your account lives on. Every post should fit one of these buckets.

Example: AI tools tutorials | Productivity hacks | Creator income | Real life + mom moments

Q5. What is your primary engagement goal?

Pick ONE. Each needs a different content style.

Q6. Which creators do you admire or compete with?

List 2–3 Instagram handles in your niche. Claude will study their patterns and find gaps you can fill.

Example: @foundr, @tanmay.bhat, @shreyasinghal

Q7. What are you currently learning or building?

Example: "Completing Anthropic's Claude 101 course, learning Notion, building a 21-day AI cohort"

This helps Claude align your content with your learning journey so you teach what you're doing — in real time.

Q8. What is your realistic time per day?

Be honest. Not aspirational.

Example: "3 hours max, split across nap time (1.5 hrs) and night (1.5 hrs). No weekends."

Q9. What is your 3-month goal?

Make it specific and measurable.

Bad: "Grow my account"

Good: "10K followers, launch one paid cohort at ₹2,999, sell 20 seats"

Q10. What is your biggest struggle right now?

Example: "I know what to teach but I don't know what to post, in what format, and how often — so I end up posting nothing"

The Full Prompt

Once you have all 10 answers ready, paste this into Claude. Fill in the brackets with your answers.

I need you to act as my full social media manager and help me plan and grow 
my presence for the next 30 days.

Before you start — ask me any clarifying questions you need to give the 
most specific, actionable output possible.

---

MY SITUATION:
- Niche/role: [your answer from Q1]
- Target audience: [your answer from Q2]
- Content tone: [your answer from Q3]
- My 3–4 content pillars: [your answer from Q4]
- Primary engagement goal: [your answer from Q5]
- Creators I admire or compete with: [your answer from Q6]
- Currently learning or building: [your answer from Q7]
- Daily time available: [your answer from Q8]
- 3-month goal: [your answer from Q9]
- Biggest struggle: [your answer from Q10]
- Platform: [e.g. Instagram — Reels + Carousels]
- Current followers: [e.g. 2,500]
- Audience language: [e.g. Hinglish / English / Hindi]
- Location/market: [e.g. India]

---

STEP 1 — MARKET RESEARCH
- Current demand and trending topics in my niche
- Content formats performing best on my platform right now
- Biggest audience pain points I can address with content

---

STEP 2 — COMPETITOR RESEARCH
- What top creators in my niche are doing well (style, frequency, hooks)
- Gaps and angles they are NOT covering that I can own
- Content types getting the most saves, shares, comments in my niche

---

STEP 3 — SEO & HASHTAG STRATEGY
- 20 high-search, low-competition keywords for my niche
- Hashtag strategy: 5 broad + 5 mid + 5 niche hashtags to rotate
- Keywords for my bio, captions, and profile name

---

STEP 4 — 30-DAY CONTENT CALENDAR
- Realistic posting schedule based on my available hours
- Reels, Carousels, and Story prompts mapped to my pillars and tone
- Content mix: educational, relatable, trending, promotional
- CTA for every post: follow / save / DM / comment / buy
- Flag posts with highest viral potential

---

STEP 5 — NOTION TRACKER
- Simple table: Content type | Topic | Status | Posted date | Engagement notes
- 5-minute weekly review habit to see what's working

---

Be specific. Be realistic for a solo creator with limited time.
No generic advice. Everything should be executable alone.

What Claude Gave Me Back

Not just a calendar.

• ✅ Market research — trending topics people are actually searching for right now

• ✅ Competitor analysis — what's working in my niche, and gaps I can own

• ✅ SEO + hashtag strategy — so people can actually find me

• ✅ 30-day Reels, Carousels, and Stories — matched to my tone and pillars

• ✅ Notion tracker — pushed and ready to use

That's a full social media manager. In one prompt.

One Thing to Remember

This works for any niche. Fitness. Finance. Education. Business. Student. Solopreneur.

You don't need to be an AI person to use this. You just need to know your own answers.

The prompt is a template. Your life is the content. Claude connects the two.

Try It and Tell Me

Copy the prompt. Fill in your details. Hit enter.

Then come tell me on Instagram what Claude planned for you 👇

📸 Follow me on Instagram: @ranjanicshetty

I post simple, practical AI tutorials for creators, students, and everyday people — no jargon, no fluff.

Ranjani Shetty is an AI educator with 10 years of ecommerce experience. She teaches working moms, students, and small business owners how to use AI tools practically — one reel at a time.

I've been using ChatGPT for close to three years. It knows how I write, what I work on, how I like my answers structured. That context took months to build up without me even realising I was building it.

When I started using Claude more seriously, that was the one thing annoying me. Not features. Not price. Just the idea of explaining myself all over again.

Claude actually has a solution for this. There's a memory import feature — you pull what ChatGPT knows about you and drop it into Claude. Here's how it works.

What you'll need

A ChatGPT account (free or paid works), a Claude account (free at claude.ai), and five minutes if you're slow about it.

Step 1 — Go to Claude's memory import page

claude.com/import-memory, or the long way: claude.ai → profile → Settings → Capabilities → Memory → Import.

Claude gives you a prompt. Don't just glance at it and move on — copy the whole thing, exactly as written.

Step 2 — Paste it into ChatGPT

New ChatGPT chat. Paste the prompt. Here's what it says, so you can do this right now without opening Claude first:

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 I've shared (name, location, job, interests). Professional context (industry, role, tools, workflows). Projects or goals I've mentioned. Preferences I've expressed about how you should behave.

What comes back is usually more than you expected. Instructions you forgot you gave it. Projects from six months ago. Preferences you set once and never thought about again.

Step 3 — Import it into Claude

Back to Claude. Settings → Capabilities → Memory → Import. Paste what ChatGPT gave you. Click Add to Memory.

Five minutes, maybe less if you're not reading every line of ChatGPT's output first (I did, it was interesting).

Check that it worked

Open a new Claude chat — not the same one — and ask: "What do you know about me?" or "What kind of work do I do?"

If it answers correctly, you're set. If something's off or missing, you can add it manually in Settings → Capabilities → Memory. Took me under a minute to fill in two things that didn't carry over.

A couple of things to know

Claude's memory is built around work context. Personal stuff that's not connected to how you actually use it may not stick — Claude prioritises what helps you work, not everything you've ever mentioned.

Also — don't quit ChatGPT. I still use both. They're good at different things, and now Claude actually knows who it's talking to.

Five minutes of setup. Weeks of not having to re-explain yourself. Worth doing once.

#Claude #ChatGPT #AITools #LearnAI #AITips #UpskillWithAI

I went through all of them so you don't have to guess where to start.

Here is exactly which course is for you.

Start Here — Everyone

No matter who you are, your first course is the same.

Claude 101

This is the starting point. It covers how Claude works, how to use it for everyday tasks, and how to prompt it well. Short, practical, no prior knowledge needed. Finish this first before anything else.

Link: https://anthropic.skilljar.com/claude-101

Never Used AI Before?

After Claude 101, do this one next.

AI Fluency: Framework and Foundations

This is not a course about AI tricks. It teaches you how to actually think about AI — when it works, when it doesn't, and how to use it responsibly. Two courses in and you will understand AI better than most people around you.

Link: https://anthropic.skilljar.com/ai-fluency-framework-foundations

Also worth doing: AI Capabilities and Limitations — a short introductory course about how AI actually works under the hood.

Link: https://anthropic.skilljar.com/ai-capabilities-and-limitations

Student?

AI Fluency for students

This one is built specifically for learning, assignments, and career planning. Not a generic course copy pasted for students — actually made for you. It covers how to use Claude for studying, how to think about AI in your career, and how to collaborate with AI responsibly.

Link: https://anthropic.skilljar.com/ai-fluency-for-students

Teacher or Educator?

You have three courses here and they build on each other.

AI Fluency for educators

How to bring AI into your teaching practice and classroom strategy. Not just what AI is — how to actually use it with students.

Link: https://anthropic.skilljar.com/ai-fluency-for-educators

Teaching AI Fluency

This one is for teaching AI Fluency to others in an instructor led setting. If you want to run AI workshops or train your colleagues, this is the one.

Link: https://anthropic.skilljar.com/teaching-ai-fluency

Three courses and you will be ahead of almost every school in India.

Work in an Office or Run a Small Business?

Introduction to Claude Cowork

This course is different from the others. It teaches you how to work alongside Claude on your actual files and real projects — not just chatting with it. If you want to save hours every week using AI in your real workflow, start here after Claude 101.

Link: https://anthropic.skilljar.com/introduction-to-claude-cowork

Work for a Nonprofit?

AI Fluency for nonprofits

This course shows nonprofit professionals how to use Claude to scale impact without hiring more people. Grant writing, donor communication, internal workflows — all covered. Practical and focused on real nonprofit problems.

Link: https://anthropic.skilljar.com/ai-fluency-for-nonprofits

Developer?

You have a full path here. Do them in this order.

Building with the Claude API — 84 lectures, over 8 hours. Covers everything you need to build real AI powered products using Claude.

Link: https://anthropic.skilljar.com/claude-with-the-anthropic-api

Introduction to Model Context Protocol — Learn to build MCP servers and clients from scratch. Connect Claude to external tools and services.

Link: https://anthropic.skilljar.com/introduction-to-model-context-protocol

Model Context Protocol: Advanced Topics — Production level MCP patterns. For when you are ready to go deeper.

Link: https://anthropic.skilljar.com/model-context-protocol-advanced-topics

Claude Code in Action — How to use Claude Code in a real development workflow.

Link: https://anthropic.skilljar.com/claude-code-in-action

Introduction to agent skills — How to build, configure and share Skills in Claude Code.

Link: https://anthropic.skilljar.com/introduction-to-agent-skills

Introduction to subagents — How to create sub-agents in Claude Code to manage complex workflows.

Link: https://anthropic.skilljar.com/introduction-to-subagents

That is a complete developer path straight from the team that built Claude.

Use AWS or Google Cloud?

Two bonus courses if you work with cloud platforms.

Claude with Amazon Bedrock — Full training on working with Claude through AWS.

Link: https://anthropic.skilljar.com/claude-in-amazon-bedrock

Claude with Google Cloud's Vertex AI — Working with Claude through Google Cloud.

Link: https://anthropic.skilljar.com/claude-with-google-vertex

Quick Summary — Your Course by Who You Are

One Last Thing

All 16 courses are at anthropic.skilljar.com — free, self paced, and you get an official Anthropic certificate when you finish.

The certificate is real and verifiable. Adding it to LinkedIn from the company that actually built Claude carries more weight than most paid bootcamp certificates I have seen.

If you want help figuring out where to start or how to use Claude in your specific work, I do 1:1 sessions at ranjanishetty.in — happy to help you find your path.

• Ranjani

I almost didn't either.

But Claude has a memory import feature that lets you pull everything ChatGPT knows about you and bring it straight into Claude. The whole thing takes under 5 minutes. I timed it.

Here's how it works.

First — Why Claude at all?

Fair question. ChatGPT is good. I still use it.

But after spending time with Claude, a few things genuinely surprised me. It handles long documents without losing the thread halfway through. It's more careful about admitting when it doesn't know something, which matters a lot if you're teaching people and don't want to pass on nonsense. And for writing and explaining things in plain language — which is basically my whole job — it just clicks.

You don't have to pick one. I use both, depending on what I need. Different tools for different jobs.

What You Need

• A ChatGPT account (free or paid, doesn't matter)

• A Claude account — free at claude.ai

• About 5 minutes

That's it.

The Steps

Step 1. Open claude.ai. Click your profile icon, go to Settings, then Capabilities, then Memory.

Step 2. You'll see an "Import Memory" section. There's a ready-made prompt there. Copy it.

Step 3. Open a new ChatGPT chat. Paste that prompt in. ChatGPT will spit out a summary of everything it knows about you — your preferences, your work, how you like responses formatted.

Step 4. Copy ChatGPT's full response.

Step 5. Back in Claude's Memory settings, click Import, paste what you copied, hit Add to Memory.

Done. Claude now knows what ChatGPT knew.

The Prompt to Paste Into ChatGPT

This comes straight from Claude's import page. Paste it exactly as written — don't edit it:

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 I've shared (name, location, job, interests).
- Professional context (industry, role, tools, workflows).
- Projects or goals I've mentioned.
- Preferences I've expressed about how you should behave.

Anthropic wrote this specifically to get the most useful output from ChatGPT. It works better if you leave it alone.

Did It Actually Work?

After importing, just ask Claude directly:

• "What do you know about me?"

• "What kind of work do I do?"

• "How do I like my responses formatted?"

If it answers correctly, you're good. If something's missing, you can always edit your memory manually in Settings — add things, remove things, whatever you want. You're in control of it.

A Few Things Worth Knowing

You don't have to quit ChatGPT. Seriously, use both. I do.

The memory import is a one-time thing. After that, Claude builds on what it learns from you in each conversation.

You don't need Claude Pro for this. The free account works.

One Last Thing

I was genuinely surprised this feature existed. Most people I've told about it had no idea — they assumed switching AI tools meant starting completely from scratch.

You don't have to.

If you found this useful, I write about AI tools plain language — no jargon, no hype. Follow along on Instagram or come back here. And if you want to figure out how to actually use AI in your work or studies, I do 1:1 sessions at ranjanishetty.in.

See you in the next one.

— Ranjani Shetty