Claude Code Users Hit With Weekly Rate Limits (techcrunch.com) 43
Anthropic will implement weekly rate limits for Claude subscribers starting August 28 to address users running its Claude Code AI programming tool continuously around the clock and to prevent account sharing violations. The new restrictions will affect Pro subscribers paying $20 monthly and Max plan subscribers paying $100 and $200 monthly, though Anthropic estimates fewer than 5% of current users will be impacted based on existing usage patterns.
Pro users will receive 40 to 80 hours of Sonnet 4 access through Claude Code weekly, while $100 Max subscribers get 140 to 280 hours of Sonnet 4 plus 15 to 35 hours of Opus 4. The $200 Max plan provides 240 to 480 hours of Sonnet 4 and 24 to 40 hours of Opus 4. Claude Code has experienced at least seven outages in the past month due to unprecedented demand.
Pro users will receive 40 to 80 hours of Sonnet 4 access through Claude Code weekly, while $100 Max subscribers get 140 to 280 hours of Sonnet 4 plus 15 to 35 hours of Opus 4. The $200 Max plan provides 240 to 480 hours of Sonnet 4 and 24 to 40 hours of Opus 4. Claude Code has experienced at least seven outages in the past month due to unprecedented demand.
AI really does move fast (Score:4, Funny)
It has already reached the enshittification stage!
Re: AI really does move fast (Score:2)
You haven't seen nothing yet. This still isn't the true cost of it.
Re: (Score:2)
This is for now a competitive market, and these quotas a pretty generous for what one license is marketed to be used for.
Re: (Score:2)
I don't agree. Charging an up-front price for a good product is not what enshittification is.
OK... that much makes sense. But is that the case here? Is the price the same and the service offering equal or better than when one first signed up? No.
That is simple price competition.
That implies that product A and product B are in competition and can compete based on price. That doesn't seem to be the case here.
Gradually reducing the value of the product by introducing ever more-annoying little profit streams is what enshittification is.
No. It's not about reducing the value of the product.
It's about degrading the quality of services, often by promoting ads and sponsored content and such, in order to increase profits. IE: adding shit that makes it worse for the
Re: (Score:2)
What I see here is quotas. And obviously, having quotas is worse for the customer than not having quotas for the same price, so it's a bit of a bummer, whatever word fits. But at least it isn't subtle or sneaky.
Re: (Score:2)
There is no such thing as a flatrate for something that has usage depending costs.
Ask Claude (Score:4, Funny)
Claude, can you write a load balancer for yourself so you don't have to impose user limits?
Re: (Score:2)
If this were about average use, they would probably not do it. This is about total load.
Re: (Score:2)
[A]I [I]nvestments
Making it a recursive acronym,( ala GNU's NOT UNIX) whose subliminal meaning is, you don't have to really care about what it actually is, just invest more in it forever, because, well because it says so.
Where it gets scary is that this now commonly available knowledge is being fed by the AI in and by itself and got caught in its own loop, demanding ever more resources for the sole purpose of demanding more funding, so it can use more resources to demand more funding
Re: Ask Claude (Score:3)
What if it's purely about creating scarcity for more subscription profit, and in reality demand is not prohibitive because, as you continually remind us, AI is no good and how can so many be enthralled with such a dumb fake?
Re: (Score:2)
Re: (Score:2)
The usual way to deal with peak-loads is load based pricing or load-based delays. Hence this would be a stupid way to address that. Not saying it is not the root-cause, but if it is they are doing it wrong.
Re: (Score:2)
How would they, as as one of many offers, create scarcity?
Blessing in disguise (Score:2)
Claude Code uses up tokens like a donkey kong machine
Re: (Score:3)
Looks like LLM use is a bit more expensive ... (Score:5, Insightful)
... than many people hallucinated. No surprise for me.
Re: Looks like LLM use is a bit more expensive ... (Score:2)
What if the price is a hallucination by the CEO?
Re: (Score:2)
In this competitive environment? Hardly.
Re: (Score:1)
I use different AI services, but never came across something like Claude. I paid for their service just to find out that the rate limit comes after about 30 minutes into using them for programming. Really Idiotic and unusuable.
Re: (Score:2)
Soo, you can do it for some homework coding and then that is it?
Better get that ... (Score:2)
Some day people will discover (Score:5, Insightful)
That they've got a pretty decent LLM built in to their own skull, and it's available gratis for their personal and professional use for up to 8-12 hours per day. The best part is that the more you use it, the better-trained and more effective if gets.
Re: (Score:1)
Re:Some day people will discover (Score:5, Informative)
That's true, for a specific and relatively narrow specialization.
However, when you need a one-off configuration, setup, solution, script, in a language you don't master, that's where having a code helper available is a godsend.
One recent case was when I wanted to have my Home Assistant home setup monitor my network devices via SNMP. What would have taken many hours, with the assorted swearing and frustration episodes, took about 3 hours total using Gemini Pro 2.5. It did all the heavy lifting for me and taught me quite a few things (with the occasional hallucination, of course, but I know hoe to weed out those occurrences).
LLMs are tools which help me be more productive. Much like knives or hammers, when used right, they are useful, and when used wrong, they are dangerous.
Re: (Score:2)
I've always figured these things really really excel as learning tools for professionals.
Re: (Score:2)
Unfortunately, my brain has a fairly strict rate limiter and I have yet to see a way to pay out of that.
Re: (Score:2)
Sigh. (Score:2)
Oh, look, now the limits start to come in, then the prices will go up, and eventually... hell... who knows... maybe the investors will actually see A PROFIT from those billions spent on training an AI to be a glorified autocomplete.
It'll only take about a hundred years or so to pay back their investment, even then.
If... (Score:4, Insightful)
Re: (Score:1)
Well,
I know about a (good no idea?) developer who had a lot of money.
He paid "online gamers" to harvest items for him in an online game.
Because he thought (and told so in public): "I love that game, and when I play it the 5h a week while I have time, I want to play it by the most potential".
Using an LLM for coding is more or less the same.
If you have to write 100 lines of code that you have clearly in your mind, and takes 3h to do right, but an LLM can spit it out in 30 seconds ... does that make you a bad
Re: (Score:2)
Re: (Score:3)
If you're willing to pay $200 per month for coding assistance, I suspect you're a shit developer.
If your'e a developer who makes $100 per hour, spending $200 per month to save even 10 minutes per day is a no-brainer. If it saves you an hour or more per day you'd have to be a complete idiot not to do it.
Re: (Score:2)
It's not "saving" you anything. You are still required to work the standard 8 hours per day like most people are required to do. With "AI" you're also required to check everything the LLM spits out for errors, which is way too often, and is also a lot of work if you don't want to create more problems for yourself later. LLMs may save you some typing with autocomplete, but generally your cognitive load with them isn'
Re: (Score:2)
It's not "saving" you anything. You are still required to work the standard 8 hours per day like most people are required to do.
If you're a contractor, you could use the savings to work less. If you're an employee, your employer would reap the rewards of your higher productivity. In either case, it's rather odd to argue that you should try to be less productive. If that's your position, you should avoid using high level languages, IDEs, debuggers, etc. Why not write your code in machine language (not assembler -- that would make you more efficient) on paper, do all of your testing with hand simulation, then use toggle switches t
Re: (Score:2)
https://www.reuters.com/busine... [reuters.com]
Re: (Score:2)
Nice try. It's been shown that using "AI" can actually slow down productive programmers, because they have to do more work to get the "AI" to produce a usable result. I find "AI" rather exhausting, honestly. Overall it's a net negative for a lot of programmers. It may not seem that way, but it is that way.
If definitely makes me faster if I use it correctly. If it slows you down, don't use it... but you may find yourself falling behind your peers who have figured out how to be productive with it.
Some things I find helpful:
Don't ask it to write your core code. Trying to explain in English what you want some complex function to do is often harder than just writing the code yourself. Definitely don't re-prompt more than a couple of times if the output isn't right -- write or fix it yourself. The goal is
Re: (Score:2)
Oh, one more suggestion (because I just did it in the other window): Do ask the LLM to make code modifications for you. Suppose you're changing a method signature in a somewhat-complex way such that you can't just search & replace, or let the IDE refactoring tool do it. Tell the LLM to find and fix all calls. Often you can be that vague, too "Find and fix all calls to my_func()". Sometimes you have to specify more precisely... but always start by telling the LLM to do it the same way you'd tell a h
Re: (Score:2)
Re: (Score:2)
You assume I have not been using "AI". I have. It sucks. I've done all the things you suggest.
Good for you, I guess.
YMMV, especially if you suck at your programming job.
Well... I'm a Staff SWE at Google with >35 years of professional software development experience, and my code running in the core OS of 3B devices. By the standards of most people, I'm a skilled, experienced and highly productive programmer. Maybe your standards are higher.
Re: (Score:2)
Since you work for Google, I've very much not impressed with the AI summary in search results either. It's bordering on a liability, even with the disclaimer that it may not be correct (i.e. slop). Nice job racing to the b
Gotta be done. (Score:2)
A company that's trying to succeed by actually making a profit instead of killing everybody else by running losses the longest will have to right-size their revenue model. A small subset of power users can really kill all profit for an AI-based service.
So long as they are applying it to new and renewing contracts, I think it's fair. Just don't try to tinker with agreements that are in flight.