As a $20 month user, I can tell you in my experience it's "refactoring" jobs that really smash through those tokens quickly. If you do a "write a component that does this" kinda thing, you can use the $20 plan almost an unlimited amount of time. If you are doing "find all instances of ComponentFoo, change to ComponentBar, refactor each screen for correct usage of ComponentBar" kinda things, it's going to grep through your code, find multiple files, read all of them into context and start making changes one by one and/or spin up a subagent to do it. You'll be rate limited pretty quick doing things that way.

Are people really using LLMs this way? Can't your IDE at least attempt global refactorings? I get that it's harder with dynamic languages like Python but using an LLM to rename a class seems like using a nuke to crack a nut.

For sure, it was a bit of a simplified example, but you can imagine prompts that are asking the LLM to touch dozens of files at once. A lot of people will use fairly high level requests, "make a screen in my app that does this" and that ends up making screens, components, queries, local state changes, etc etc. It adds up quick having so much in the context at once (and if you /clear the context and ask it to do anything more than one off changes in a file, it's gonna grep and glob it's way through your code looking for the right files)

That kind of gives a point in favor of AI enabled IDEs, these kind of operations could often be achieved with simple refactoring tools available without AI today and save massive amount of trees.

> If you are doing "find all instances of ComponentFoo, change to ComponentBar, refactor each screen for correct usage of ComponentBar" kinda things, it's going to grep through your code, find multiple files, read all of them into context and start making changes one by one and/or spin up a subagent to do it. You'll be rate limited pretty quick doing things that way.

Huh?? grep and sed do this for free; you don't need A.I. for that.

For sure, Claude Code specifically has a grep and glob tool (as well as access to bash). The LLM will decide to use those tools to find files, then once it has found matches to what it's looking for, it reads those files into its context. It will then decide if it indeed was a successful match and start working on that file. There's a lot of times it finds a match using grep, reads that into its context and then decides "no, that's not what I was actually looking for..." and moves on.

Unlike some other AI coding IDEs, Claude Code doesn't either keep your entire codebase in context, or use some sort of vector representation and vector search of your code. It basically uses grep to find a keyword it's looking for then reads the file (it can also just read a few lines of a file too) into its context.

It seems to be the "magic" of Claude Code..it's very UNIX like, uses existing tools to do what it needs to do instead of trying to re-invent the wheel, so to speak.

$20/mo plan doesn't include opus (the larger model) like the $100+ plans do, it's likely they are hitting the opus limit which is fairly low.

The $20/month plan most definitely does include Opus. Just not a ton of usage allowed.

Only opus on web. Not Claude code

I'm on the $20 and was hitting the limit quite often. It lasted 2.5 h out of 5 h if I went all in. So even before this questionable change, we were at 50% utilization at most.

And I guess it'll go downhill from here. Anthropic, I wish you the best. Claude is a great tool at good value. But if you keep changing the product after my purchase, that's bad value.