[rsanek]

I was convinced of this line of thinking for a while too but lately I'm not so sure. In software in particular, I think it's actually quite relevant what you can do in-house with a SOTA model (especially in the tool calling / fine tuning phase) that you just don't get with the same model via API. Think Cursor vs. Claude Code -- you can use the same model in Cursor, but the experience with CC is far and away better.

I think of it a bit like the Windows vs. macOS comparison. Obviously there will be many players that will build their own scaffolding around open or API-based models. But there is still a significant benefit to a single company being able to build both the model itself as well as the scaffolding and offering it as a unit.

[mritchie712]

CC being better than Cursor didn't make sense to me until I realized Anthropic trains[0] it's models to use it's own built-in tools[1].

0 - https://x.com/thisritchie/status/1944038132665454841

1- https://docs.anthropic.com/en/docs/agents-and-tools/tool-use...

[jgraettinger1]

It still doesn't make sense. Cursor undoubtedly has smart engineers who could implement the Anthropic text editing tool interface in their IDE. Why not just do that for one of your most important LLM integrations?

[mritchie712]

I agree it doesn't make sense. I'd think they could alias their own tools to match Anthropic's, but my guess is they don't want to customize too heavily on any given model.

[klausa]

My gut feeling is that Claude Code being so popular is: - 60% just having a much better UX and having any amount of "taste", compared to Cursor - 39,9% being able to subsidize the raw token costs compared to what's being billed to Cursor - 0,1% some magical advantage by also training the model

Claude Code is just much _pleasant_ to use than most other tools, and I think people are overly discounting that aspect of it.

I'd rather use CC with slightly dumber model, than Cursor with a slightly better one; and I suspect I'm far from being the only one.