Why do Cursor, Windsurf and Claude Code dominate the conversation?

Perhaps a combination of the Matthew and the network effect. https://en.wikipedia.org/wiki/Matthew_effect https://en.wikipedia.org/wiki/Network_effect

Why do Pepsi and Coke dominate the cola conversation?

Because they are the most popular.

I was surprised recently to see that gemini-cli[0] and codex[1] each have way more GitHub stars than Claude Code[3]. Currently 62k, 31k, and 25k respectively.

Stars are only a proxy for use, of course, but I'm not sure what a closer public indicator might be.

1. https://github.com/google-gemini/gemini-cli

2. https://github.com/openai/codex

3. https://github.com/anthropics/claude-code

There is basically no reason to use a tool that is "almost as good" for the same price as the best one. So people concentrate on the best ones out there.

There is a YouTuber, GosuCoder [0], who is benchmarking all of the agentic coder tools and in some of his recent videos you will see Claude Code is currently ahead of the field.

[0]: https://youtu.be/Lon0oRRqB6A?t=334

Three is already a lot of tools to be doing more or less the same thing. Not many people are going to have informed, experience-driven opinions on Cursor, Windsurf, and Claude Code, let alone the other tools you mention.

As boring an answer as it may be, mindshare matters. Newcomers face outsized inertia in winning that mindshare over from the major incumbents.

For Replit, bolt, lovable and similar tools are very frustrating to work as a software engineer.

V0 is talked about a bit more but they did a good job specializing it more on frontend (not trying to do everything)

Maybe that is why in circles dominated by people closer to the code, those tools aren’t discussed much.

From my perspective, Gemini CLI is useless—it’s the only LLM that has repeatedly refused to work with me because of so-called "offensive language." And when it does work it's overly verbose and often just plainly wrong. Codex CLI just feels like a stripped-down version of Claude Code.

Claude Code wins for me for one simple reason: I can just pay for a MAX account and not worry about running out of credits or going through my credit card limit. That makes me use the tool twice as much as the others, where I’m constantly watching my credit balance.

Cursor wins because it was the first big one, not sure if it has any important advantages over claude code anymore.

Why do worse products win out or have more mindshare over better products? It is an old question, and the answer is it depends. Use what you think is best, and tell others.

As a CC user I can say there there are at least two reasons why CC dominates.

One, it is very good except for its prise. Because of the cost, I tried Codex, but after a few simple questions I gave up. Nowhere close to CC when it comes to tool usage. Gemini cli asked for a Cloud project id, no idea what is meant with this. Cursor is not an option because I think long term they can’t compete with the integration that Anthropic can offer.

Second, CC offers capabilities you will learn as power user, in that sense it is like a game; addictive.

I use Sourcegraph Amp (plugin for vscode) and am very happy with it. I assume it works like all the others. I experimented a bit with CC and Cursor (some months ago) but for me Amp works more easy.

I do wonder why never ever anybody else mentions Amp though. Am I their only user?! It explains why it works so fast ;)

I’ve used Claude and Gemini and Claude works a lot better, but Gemini is a bit cheaper so I have both and use Gemini when I don’t want to waste my Claude availability.

However, Claude wastes less of my time than Gemini.

In my experience, Gemini is a lot more likely to have an empty tool call and get in a loop of using a tool that isn't working, or get confused and use some previous instruction from the context window instead of the task at hand and go on a totally crazy tangent. Claude is generally a lot more accurate.

I recently tried Replit. After it chewed through all the available free agent credits. It gave up. I then received an email that my account was banned for violating the terms of service (which includes all sorts of squishy language).

Replit is a cash grab. It couldn't build a basic app with Three.js without stepping on itself.

Claude Code is the only AI tool that can grok my C++ codebase.

Cursor feels right so long as you have proper .md's

Waiting for access to Kiro to try it.

Currently running GGUF models on my own with an OpenAI mock server I wrote so Continue.dev works in VS Code as an agent. Meh. I don't want to spend a bunch of money for a project that I can't use in the end so I don't waste my money on tokens.

It’s very simple. They’re the only good products.

You can’t underestimate the viral nature of vibe coding which was almost exclusively done using Cursor and Windsurf. There are tons of YouTube videos boosting both products.

Claude Code grabbed tons of attention because it uses the command line and enables developers to boost their productivity.

Because it works via the command line, it doesn’t require using an editor/IDE you may not use. It’s extensible via custom slash commands and hooks.

It’s only been out a few months and there’s already a cottage industry of websites, tutorials and videos supporting it.

It’s not perfect but it’s really good.

I’ve never seen so many developers making videos and blogposts describing how happy they are to pay $200/month for the Max plan due to the productivity gains.