You have seen the make money with AI gurus. They sell a course for hundreds of dollars. Some charge $499. I looked hard at what is really inside them. It is thin. Every real step is free, and it is all on this page.
I am not selling you a dream. I build real apps with AI. Two of mine are live on the App Store and the Play Store right now. I also burned over two thousand dollars learning this the slow way, so you do not have to. This is the exact playbook I use to build things people pay for.
This is my real playbook, free. The real steps, the real prices, and the honest warnings the course sellers skip. You will not find it written down anywhere else. And if any part looks too hard, relax. At the end I hand you two prompts that do the heavy lifting for you.
5 honest truths first
Read these once. They saved me months of wasted work. If a guide skips them, it is selling you a dream.
- Most people who try this make nothing. Not because AI cannot build. Because they build something generic, or they never ship. This guide kills both.
- Fast code is not safe code. Studies found nearly half of AI written code ships with security holes. One fully AI coded app leaked around 1.5 million API keys. You get the safety checks later.
- You are the director, not the audience. Vague prompt in, useless app out. Your thinking is the product. The code is just the output.
- First dollar beats first million. Forget $10,000 a month. Get one stranger to pay you one dollar. That single sale is the only proof that matters.
- This is a skill, not a lottery. You will do it badly once, okay the second time, then it clicks. That is normal. Keep going.
The 5 ways to actually get paid
Pick one. Trying all five means doing none. Choose the one that fits you today, not all of them at once.
- Build it for them. The fastest cash. Make small tools or sites for local businesses and creators. Money this week, no audience needed.
- Micro SaaS. My recommended start. A small web app that solves one problem, charged monthly. You own it, and it pays while you sleep.
- Sell products in your own store. Sell digital or physical goods yourself. You keep more, since stores take nothing on goods you sell on your own site.
- Native phone app. A real app on the stores. The most gatekeeping path. Fees, review, and testing. Full reality below.
- Templates and digital products. Build once, sell it many times. Slow to start, passive once it lands.
My honest pick for you: start with option 1 or option 2. I went straight for the app stores and hit every wall, the fees, the review, the testing. Start smaller. Earn that store path later, after you ship something and get your first paying user.
The core loop: idea to first dollar
Six moves, from setup to your first sale. The same six whether you build a site, a SaaS, or a store. Memorize this loop: idea, then PRD, then build, then ship, then sell.
Step 0: Set up Claude Code
Claude Code is the AI that builds your app. It comes with a paid Claude plan. Pro is about $20 a month and includes it. Heavy builders move up to Max, about $100 a month. That $100 is what I pay. You do not need a powerful laptop.
Claude plans and current pricesNot sure how to install it? I wrote a full step by step setup guide for non coders. Do that one first, then come back here. It reassures you the whole way.
My full Claude Code setup guide (do this first)Step 1: Pick a real problem
An app is not an idea. Be painfully specific. Not a photo app. A tool that lets wedding photographers send a private gallery and get paid before they deliver. Niche, specific, and someone already wastes time or money on it. Generic is dead on arrival.
Stuck on an idea? Look at your own job or hobby. What boring task eats your time every week? That is your first product. You already know the problem better than any guru. This is the part I spend the most time on, and so should you.
Step 2: Turn it into a PRD
A PRD is a product requirements doc. It is just a clear one page plan for your app. This is the clear intent that separates money from mush. Do not skip it. My first prompt below writes it with you, so you never stare at a blank page.
Step 3: Build it in stages
Never say build the whole thing. Paste your PRD, then build in small reviewable stages. First the layout and one real screen. Then the database and security. Then the main feature. Then payments.
Why stages? Reviewing each stage is how you catch mistakes early and keep control. Build it all is how you end up with a broken box you cannot fix. My second prompt below runs these stages for you, so you stay in charge.
Step 4: Put it live for $0
Host it free on Vercel. The free tier runs real apps. Use Supabase for the database and login, also free to start. Take payments with Stripe. Stripe has no monthly fee, just a small percent per sale. You only pay once you actually earn.
Vercel (put your app online free)Supabase (database and login)Stripe (take the payments)Stop and look at what you just did. Your app is now live on the real internet, for zero dollars up front. That is the part most people never reach. The hard part is behind you.
Step 5: Get your first customer
This is the only step that counts. Post where your niche hangs out. DM ten people who have the problem. Give it free to three for feedback. Charge the fourth. One paying stranger is proof, and it is the moment this stops being a hobby. Then improve and repeat.
The two prompts that do the heavy lifting
Here is the real sauce, the part the gurus charge you for. Two prompts I built from shipping my own apps. The first writes your plan with you. The second builds your app in safe stages. Paste them into Claude Code exactly as written.
This part has a little structure in it. You do not need to understand every word. Just paste it and answer the questions Claude asks you. It is easier than it looks.
you are my product co-founder. here's what i want to build. [one sentence, what it does and who it's for] ask me 5 sharp questions to nail down the target user, the single core problem, and the one feature that must work. then write a concise prd that covers the problem, the target user, the core user flow step by step, the must-have features (max 5), the data to store, and what's out of scope for v1. keep it under one page.
You should now have a one page PRD. That is your blueprint. The more you sharpen it, the better your app comes out. Save it, you will use it in the next prompt.
here's my prd. [paste it] build this as a next.js (app router) app with shadcn/ui and supabase (auth and database), ready to deploy on vercel. work in stages and stop for my review after each one. 1) scaffold plus one working core screen 2) database schema plus row level security on user data 3) the core feature, end to end 4) stripe checkout for payments do not touch production credentials. explain each step in plain english as you go.
Claude now builds your app one stage at a time and stops for you to look. You are directing a builder, not writing code. This is exactly how I build the things I sell.
Want it on the app store? Read this first
A web app skips all of this. No fees, no review, no waiting. Go native only if your idea truly needs a phone. Here is the real cost if it does, from someone who has shipped to both stores.
- Apple: $99 a year. Review usually takes 1 to 3 days. The app must be complete and polished, with no demo feel.
- Google: $25 one time. Cheaper, but a new personal account must run a 12 tester, 14 day closed test before going live, plus ID checks. Budget 2 to 4 weeks.
- Store fees: 15% on digital sales under $1 million a year, 30% above. Free apps and physical goods pay nothing.
So start with a web app unless the idea genuinely needs to live on a phone. I earned the store path the hard way. Earn it later, after your first web app has real paying users.
Your first 7 days
A week is enough to go from nothing to your first dollar. You will not do it perfectly. Do it anyway. Week two is where it gets good.
- Day 1: Set up Claude Code. Pick one niche problem. Write it in one sentence.
- Day 2: Write the PRD with prompt 1. Answer its five questions honestly.
- Day 3 to 4: Build the app in stages with prompt 2. Review each stage.
- Day 5: Deploy to Vercel. It is officially live on the internet.
- Day 6: Add Stripe payments. Ask Claude to scan for security holes before you deploy.
- Day 7: Show ten real people. Get your first paying customer. The only metric that counts.
Done and live beats perfect and imagined. Every day you do not ship, someone less talented ships instead, and learns faster than you. I say that as the guy who waited too long once.
The 5 mistakes that keep people at $0
- Never shipping. A live, ugly app beats a perfect idea in your head. Ship in days, not months.
- Building generic. If it could be any app, it is no app. The narrower the niche, the easier the sale.
- Ignoring security. Turn on Supabase row level security. Never put secret keys in the app front end. Ask Claude to scan for security issues before you deploy.
- Giving Claude the keys to everything. Do not let it run loose on your live database. Review each stage. Keep your live app separate.
- Not talking to users. The app is not the business. The customer is. Talk to ten people before you build, and ten after.
If you get stuck
- Claude built too much at once and it broke: tell it to work in stages and stop after each one. Fix the real cause, not the symptom.
- Your app will not run: paste the error to Claude and say fix the real cause, not a quick patch.
- You are worried about security: ask Claude to scan the app for security holes, and turn on Supabase row level security.
- No one is buying: your niche is too broad, or you have not asked enough people. DM ten more who have the exact problem.
- Your Claude bill looks high: make sure you did not set an ANTHROPIC_API_KEY. Your paid plan already covers Claude Code.
- Do I need to know how to code?
- No. You describe what you want in plain English. Claude writes the code. My two prompts above do the hard parts for you.
- How much does this cost to start?
- You can start for about $20 a month for Claude Pro. Vercel, Supabase, and Stripe are free until you earn. Stripe takes a small percent only when you get paid.
- How fast can I make my first dollar?
- Many people go from idea to a live app in about a week. Your first paying customer depends on how narrow your idea is and how many people you ask.
- Do I have to build a phone app?
- No, and you should not start there. A web app skips the store fees, the review, and the 14 day test. Go native only if your idea truly needs a phone.
- Are the make money with AI courses worth it?
- Almost never. Most sell you a workflow you can get free, like this one. If someone charges you hundreds of dollars for it, that is the whole con.
- Is AI built code safe to sell?
- Only if you check it. Studies found nearly half of AI written code has security holes. Turn on Supabase security and ask Claude to scan before you ship.
- This still feels too advanced. What do I do?
- Relax, this is the part I promised you. Start with my free setup guide, install Claude Code, then paste the two prompts. They do the heavy lifting and explain each step in plain words.
This free guide is the what. If you want the how, done with you, so you ship your first product fast, that is my Claude for Beginners course. Want agents, apps, and automations all in one place? That is All-Access. And if you want me to build your first money making app with you live, I take a few 1:1 people each month. You are not chasing a make money scheme now. You are building a real product people pay for.