Top
Best
New

Posted by y42 15 hours ago

I cancelled Claude: Token issues, declining quality, and poor support(nickyreinert.de)
849 points | 500 commentspage 5
duxup 12 hours ago|
I’ve definitely encountered a drop in Claude quality.

Even a simple prompt focused on two files I told Claude to do a thing to file A and not change file B (we were using it as a reference).

Claude’s plan was to not touch file B.

First thing it did was alter file B. Astonishing simple task and total failure.

It was all of one prompt, simple task, it failed outright.

I also had it declare that some function did not have a default value and then explain what the fun does and how it defaults to a specific value….

Fundamentally absurd failures that have seriously impacted my level of trust with Claude.

isjcjwjdkwjxk 15 hours ago||
Oh no, the unreliable product people pretend is the next coming of Jesus turned out to be thoroughly unreliable. Who coulda thunk it.
nikolay 14 hours ago||
I can agree. ChatGPT 5.5 made this a no-brainer choice. Anthropic are idiots removing Claude Code from the Pro plan. They need to ask Claude if what they did was a natural intelligence bug! Greed kills companies, too!
Capricorn2481 14 hours ago||
> I can agree. ChatGPT 5.5 made this a no-brainer choice

The new model that came out less than 24 hours ago made this obvious? This feels like when a new video game comes out and there's 1,000 steam reviews glazing it in the first hours of release. Don't you think you should use it for longer than a day before declaring it a game changer?

nikolay 13 hours ago|||
Plus, I don't like Anthropic allowing "chosen people" to access models - OpenAI would offer anything to anybody who's willing to pay for it, and that's a real business. Anthropic claims some moral superiority, but their model is used to kill civilians in Iran and to spy on all of us, so I don't buy their BS!
nikolay 13 hours ago|||
Yes, even 5.4 was better.
robotnikman 11 hours ago||
>removing Claude Code from the Pro plan

Wait really? I wanted to give it a try, but for $200 a month no way am I paying that for something I just want to experiment around with

nikolay 11 hours ago|||
It was all over the Hacker News, but they do have a $100/mo Max plan. But so does OpenAI now, too. I guess companies assume people are willing to cough up hundreds of dollars per month, and we are. I need to review all my AI subscriptions, as it's over $1,000 per month now - $250 for Gemini Ultra (although it comes with tons and tons of other benefits), plus, it could be shared with my family - other companies live in a world where families don't exist, and people don't share a work laptop for both personal and corporate stuff. Add $200 for Claude Max and $100 for ChatGPT Pro. Add Cursor, add Zed, add Lovable, add a bunch of other things I'm experimenting with... It's getting expensive!
nikolay 11 hours ago|||
Let me take this back - you are probably grandfathered into having Claude Code in your Pro plan though! Now I feel sorry I upgraded to Max as I could've kept Pro, just in case, but now I will just have to cancel it if I want to move to ChatGPT, or Gemini 4 when it's out, and if it's worth it. They need to consult with AI when making such stupid choices, honestly!
throwaway2027 15 hours ago||
Same. I think one of the issues is that Claude reached a treshold where I could just rely on it being good and having to manually fix it up less and less and other models hadn't reached that point yet so I was aware of that and knew I had to fix things up or do a second pass or more. Other providers also move you to a worse model after you run out which is key in setting expectation as well. Developers knew that that was the trade-off.

I think even with the worse limits people still hated it but when you start to either on purpose or inadvertently make the model dumber that's when there's really no purpose to keep using Claude anymore.

dostick 12 hours ago||
The discussion about Claude always omit the important context - which language/platform you’re using it for. It is best trained for web languages and has most up to date knowledge for that. If you use it for Swift it is trained on whole landfill of code and that gives you strong bias towards pre-Swift 6 coding output. Imagine you would give Claude a requirements for a web app, and it implements it all in JQuery. That’s what happens with other platforms.
adamors 12 hours ago|
It’s not ommited, OP clearly talks about editing Javascript.
rurban 11 hours ago||
That's bad for him, because he already had a cheap plan. Now he wont get it back that easy.

Pro is gone. OpenAI plans are more expensive. He can only buy a Kimi plan, which is at least better than Sonnet. But frontier for cheap is gone. Even copilot business plans are getting very expensive soon, also switching to API usage only.

kx_x 11 hours ago||
After the fixes in Claude Code, Opus 4.6/4.7 have been performing well.

Before the fixes, they were complete trash and I was ready to cancel this month.

Now, I'm feeling like the AI wars are back -- GPT 5.5 and Opus 4.7 are both really good. I'm no longer feeling like we're using nerfed models (knock on wood)!

lawrence1 13 hours ago||
The timeline of the first few sentences doesn't add up. how can you subscribe 2 weeks ago when the problem started 3 weeks ago.
y42 10 hours ago|
https://news.ycombinator.com/item?id=47894155

(I am just learning that "a couple of weeks" apparently means "2 weeks"...)

r0fl 5 hours ago||
I hope codex doesn’t decline the same way

I’m blown away by how good it is lately

exabrial 12 hours ago|
It's bad, really bad.

The filesystem tool cannot edit xml files with <name></name> elements in it

More comments...