Hey Reader,

I just spent hours building the exact same app twice.

Same prompt. Same tech stack. Same features.

The only difference was that one used Claude Sonnet 4.5 (Claude Code), and the other used GPT-5 (Codex CLI).

And holy sh*t, the results were not even close.

Let me show you what happened.

The Test

I built a simple to-do list app with user authentication.

Nothing crazy. Just login, add tasks, edit them, delete them.

I even told both models: "Keep it super secure."

Should be easy enough, right?

Wrong.

Round 1: Sonnet 4.5

Sonnet starts fast. Like, ridiculously fast.

About 2.5 minutes later it said: "Cool, here's your app!"

Except... it wasn't working.

15 errors. Then 20. Then more.

Each time I paste an error, Sonnet says "Oh yeah, here's the fix!"

But it's not actually fixing anything.

It's creating NEW problems while "solving" the old ones.

It's like you talk to a Fiverr dev who just doesn't get you.

After 30 minutes, the app kind of works.

But then I notice something weird.

The entire to-do interface is visible... even when I'm NOT logged in.

There's a sign-in button at the top, but the app is already showing.

So I ask: "Is this secure?"

Sonnet's response: "Nope. We're using weak security that can easily be broken"

Bro. YOU built this. 😭

Round 2: GPT-5

Same exact prompt. Same tech stack.

GPT-5 goes quiet for 18 minutes.

I'm thinking: "Is this thing even working?"

Then it comes back with the entire app. There's minor errors.

Another 12 minutes pass, but then...

Working. Secure. Done.

Zero other errors.

I ask it the same question: "Is this secure?"

GPT-5: "Yes. We're already using the best practices with proper server-side checks."

It caught the security flaw that Sonnet CREATED.

What This Actually Means

Here's the uncomfortable truth nobody talks about:

Sonnet is fast but doesn't think enough.

It takes 5 seconds to "understand" a complex feature.

GPT-5 takes 8-10 minutes. Because it's actually thinking.

Sonnet gets stuck in debug loops.

You paste an error. It "fixes" it. Creates a new error. Repeat forever.

GPT-5 actually debugs. It finds the root cause.

Sonnet has security blind spots.

It stored a critical secure piece of data in a place where any hacker could exploit it.

GPT-5 caught it without me even asking from the start.

So Which One Should You Use?

Here's my honest take:

Use Sonnet 4.5 if:

• You're learning and want to iterate fast
• You like being hands-on every 30 seconds
• You're building small, simple features
• You have a Claude Max subscription and don't want to switch

Use GPT-5 if:

• You're building complex features in large codebases
• You want the AI to work autonomously
• You care about security (you should)
• You don't mind waiting for superior results

Think of it like this:

Sonnet is your intern. Fast, eager, needs constant supervision.

GPT-5 is your senior developer. Slower, but you can trust it to work alone.

My Recommendation

For my AI Coding Blueprint students, I now recommend starting with GPT-5.

Yes, it's slower.

But you'll spend way less time debugging. Way less time going in circles.

And you'll actually learn secure practices instead of creating security holes.

Speed means nothing if you're building the wrong thing fast.

Talk soon,
Rob

P.S. Want to work with me? I'm reopening my blueprint next month but you can join the waitlist.