Why I Don't See The Hype in Claude Code

This article is not sponsored btw!

Could there be something I'm urgently missing?

From my LinkedIn feed, YouTube, and X, I see the virality of Claude Code everywhere, but after having used it a couple of times, I struggle to see the hype, especially when compared to Cursor.

"I vibe-coded an app in a weekend, and now it makes $100K MRR"

Is, if you're a Twitter/X user, a line we're all too used to seeing.

Claude Code is made out to be the engine that drives results like this (well, if they're genuine).

But—1. Results like this existed before Claude Code.

And, 2. Athropic's models are not the only LLMs in the world that are capable of building apps.

Why is Claude Code getting so much of the credit?

In my opinion, it's less of the technical breakdown, but rather the positioning. 'Claude Code' is a buzzword, a narrative, a label. A neat shareable way to complement the '$XRR in 2 weeks' post with a 'hacker'/terminal aesthetic.

Why this is the case isn't entirely clear. Maybe because of its portability, 'fashionability' in the dev space, and arguably, it being the first true LLM CLI tool (although that is disputable).

The best tools don't necessarily always win—the most talkable/viral ones often do.

I prefer Cursor's developer experience

This is one of those 'real devs use Neovim' moments.

I don't buy into the pressure that the more terminal-heavy your workflow is, the 'stronger' an engineer you are.

I like using a UI to manage my LLMs. I like being able to manage agents in a visual list that is tangible, without executing a dozen commands on a black and green terminal. What is so wrong with this?

An engineer is simply a problem solver. If a local auto repair garage wants a website, an engineer who spins up a clean WordPress site in a day over the guy who writes 3 microservices deployed on Kubernetes is often the one who provides more value in most cases.

Composer 2.0 is insanely fast

Composer 1.5 was insanely fast, but 2.0 is outrageous; it almost feels like cheating. And none of the 'garnishing ✨', 'musing ✨' flowery theatrics (lol).

If I am completely honest, and I am happy to be proven wrong, I have not worked with a single LLM that is as fast for development and token-efficient as Composer 2.0.

TypeScript errors are notoriously verbose and convoluted (see this github post). These can quickly turn into a timesink. I find myself simply wanting to just click a button and forget about it, instead of asking "what's causing the lint error" several times, something that is easily doable in Cursor quickly.

It's cheaper (albeit marginally)

For basic individual plans, they are as follows:

Cursor: $20 / month

Claude Code: also $20 / month

But when it comes to annual plans, Cursor is slightly cheaper at $16, vs $17 for Claude Code.

That being said, Cursor provides access to dozens of models while Claude only supports Anthropic models, with both being the exact same monthly price. Just another reason I am unable to justify the move-over.

A diverse LLM garden is always better than a single proprietary lane, in my opinion.

Happy to be proven wrong, but I cannot see the hype

So, I am getting a model (Composer 2.0) that does implementation tasks supremely fast with reasonable token usage, a ton of other LLMs to choose from, and a GUI that is intuitive. All for the same price as a CLI tool that only supports its own vendor's models (no shade to Anthropic, btw)? That is value in my opinion.

The TL;DR from me is: Both are powerful tools, but the hype in Claude Code is something I'm unable to see right now, beyond '🤯🔥' emojis and questionable Twitter/X 'MRR' posts.

This article was not sponsored (as mentioned before).