Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I cancelled Claude: Token issues, declining quality, and poor support

nickyreinert.de

Post reply on HN
Print view View original

381–390 of 604 posts

Re: I cancelled Claude: Token issues, declining quality, and poor support

#381

Post by vivin »

This is interesting to me, because Claude has been a net-positive for me. I haven't faced token issues or declining quality. I generally work with Claude as an assistant -- I may have it do planning and have it "one shot" a thing, but it's usually a personal tool or a utility that I want it to write.

For actual code that goes out to production, I generally figure out how I would solve the problem myself (but will use Claude to bounce ideas and approaches -- or as a search engine) and then have Claude do the boring bits.

Recently I had to migrate a rules-engine into an FSM-based engine. I already had my plan and approach. I had Claude do the boring bits while I implemented the engine myself. I find that Claude does best when you give it small, focused, incremental tasks.

Re: I cancelled Claude: Token issues, declining quality, and poor support

#382

Post by root_axis »

ac29 wrote: post #186
root_axis wrote: post #87

Earlier quoted context omitted.

Not really. The hardware requirements remain indefinitely out of reach. Yes, it's possible to run tiny quantized models, but you're working with extremely small context windows and tons of hallucinations. It's fun to play with them, but they're not at all practical.

The memory requirements aren't that intense. You can run useful (not frontier) models on a $2-5K machine at reasonable speeds. The capabilities of Qwen3.6 27B or 35B-A3B are dramatically better than what was available even a few months ago. Practical? Maybe not (unless you highly value privacy) because you can get better models and better performance with cheap API access or even cheaper subscriptions. As you said, t…

> The capabilities of Qwen3.6 27B or 35B-A3B are dramatically better than what was available even a few months ago.

Yes, a lot better, but still terribly unreliable and far less capable than the big unquantized models.

Re: I cancelled Claude: Token issues, declining quality, and poor support

#383

Post by xpe »

rsanek wrote: post #378
wg0 wrote: post #202

I write detailed specs. Multifile with example code. In markdown. Then hand over to Claude Sonnet. With hard requirements listed, I found out that the generated code missed requirements, had duplicate code or even unnecessary code wrangling data (mapping objects into new objects of narrower types when won't be needed) along with tests that fake and work around to pass. So turns out that I'm not writing code but I'm r…

I'm confused. If you have detailed, specific expectations, why aren't using the best model available? Even if you were using Opus 4.7, I would inquire if you're using high/xhigh effort by default. Feels crazy to me for people to use anything other than the best available.

I also have the same question. That said, for some problems, at least over the last week or so, I did sometimes get better results from lower-effort Opus or even Sonnet. Sometimes I get (admittedly this is by feels) a better experience from voice mode which uses Haiku. This is somewhat surprising in some ways but maybe not in others. Some possible explanations include: (a) bugs relating to Anthropic's recent post-mortem [1] or (b) a tendency for a more loquacious Claude to get off in the weeds rather than offering a concise answer which invite short back-and-forth conversations and iteration.

[1]: https://www.anthropic.com/engineering/april-23-postmortem ... but also see the September 2025 one at https://www.anthropic.com/engineering/a-postmortem-of-three-...

Re: I cancelled Claude: Token issues, declining quality, and poor support

#384

Post by subscribed »

taytus wrote: post #179
ramoz wrote: post #109

Earlier quoted context omitted.

What provider do you use for Kimi

Straight from them, but I know other providers like io.net can be faster but I like to directly support the project.

Thx. I'll try with my personal projects (because dues to the data collection and ToS most providers are forbidden in my company), if I can opt out of training on my input.

I'm just getting a but tired of using Opus 2.6 which eats my whole allowance and then some £££ going through the 4kB prompt to review ~13 kB text file twice - and that's on top of the sometimes utter bonkers, bad, lazy answers I'm not getting even from the local Gemma 4 E4B.

Re: I cancelled Claude: Token issues, declining quality, and poor support

#385

Post by taikahessu »

ragequittah wrote: post #369
taikahessu wrote: post #343

Earlier quoted context omitted.

In order for LLM to be useful, you need to copy and steal all of the work. Yes, you can argue you don't need the whole work, but that's what they took and feed it in. And they are making money off of other people's work. Sure, you can use mental jiujutsu to make it fair use. But fair use for LLMs means you basically copy the whole thing. All of it. It sounds more like a total use to me. I hope the free market and tec…

I always wonder if anyone out there thinks they're not making money off of other people's work. If you're coding, writing a fantasy novel, taking a photograph or drawing a picture from first principals you came up with yourself I applaud you though.

You are absolutely right.

Seriously though, I do think that is the case. It would be self-righteous to argue otherwise. It's just the scale and the nature of this, that makes it so repulsive. For my taste, copying something without permission, is stealing. I don't care what a judge somewhere thinks of it. Using someone's good will for profit is disgusting. And I hope we all get to profit from it someday, not just a select few. But that is just my opinion.

Re: I cancelled Claude: Token issues, declining quality, and poor support

#386

Post by lp0_on_fire »

rsanek wrote: post #378
wg0 wrote: post #202

I write detailed specs. Multifile with example code. In markdown. Then hand over to Claude Sonnet. With hard requirements listed, I found out that the generated code missed requirements, had duplicate code or even unnecessary code wrangling data (mapping objects into new objects of narrower types when won't be needed) along with tests that fake and work around to pass. So turns out that I'm not writing code but I'm r…

I'm confused. If you have detailed, specific expectations, why aren't using the best model available? Even if you were using Opus 4.7, I would inquire if you're using high/xhigh effort by default. Feels crazy to me for people to use anything other than the best available.

> Feels crazy to me for people to use anything other than the best available.

Not everyone has unlimited budgets to burn on tokens.

Re: I cancelled Claude: Token issues, declining quality, and poor support

#387

Post by y42 »

lawrence1 wrote: post #210

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.

https://news.ycombinator.com/item?id=47894155

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

Re: I cancelled Claude: Token issues, declining quality, and poor support

#388

Post by joquarky »

binaryturtle wrote: post #85

I have a simple rule: I won't pay for that stuff. First they steal all my work to feed into those models, afterwards I shall pay for it? No way! I use AI, but only what is free-of-charge, and if that doesn't cut it, I just do it like in the good old times, by using my own brain.

[flagged]

Re: I cancelled Claude: Token issues, declining quality, and poor support

#389

Post by throwaway7783 »

wg0 wrote: post #262
Aurornis wrote: post #232

Earlier quoted context omitted.

Writing detailed specs and then giving them to an AI is not the optimal way to work with AI. That's vibecoding with an extra documentation step. Also, Sonnet is not the model you'd want to use if you want to minimize cleanup. Use the best available model at the time if you want to attempt this, but even those won't vibecode everything perfectly for you. This is the reality of AI, but at least try to use the right mod…

> Writing detailed specs and then giving them to an AI is not the optimal way to work with AI. It is NOT the way to work with humans basically because most software engineers I worked with in my career were incredibly smart and were damn good at identifying edge cases and weird scenarios even when they were not told and the domain wasn't theirs to begin with. You didn't need to write lengthy several page long Jira ti…

This is not true in my experience at all. I never write such detailed spec for AI - and that is my value as the human in the loop - to be iterative, to steer and make decisions. The AI in fact catches more edge cases than I do, and can point me to things that I never considered myself. Our productivity has increased manyfold, and code quality has increased significantly because writing tests is no longer a chore or an afterthought, or the biggest one for us - "test setup is too complicated". All of that is gone. And it is showing in a decrease in customer reported issues

Re: I cancelled Claude: Token issues, declining quality, and poor support

#390

Post by hirvi74 »

wg0 wrote: post #202

I write detailed specs. Multifile with example code. In markdown. Then hand over to Claude Sonnet. With hard requirements listed, I found out that the generated code missed requirements, had duplicate code or even unnecessary code wrangling data (mapping objects into new objects of narrower types when won't be needed) along with tests that fake and work around to pass. So turns out that I'm not writing code but I'm r…

That is why I still use the Chatbots and not the CLI/desktop tools. I am in 100% control. I mainly ask question surrounding syntax with languages I am not well experienced in, snippets/examples, and sometimes feedback on certain bits of logic.

I feel like I have easily multiplied my productivity because I do not really have to read more than a single chat response at a time, and I am still familiar with everything in my apps because I wrote everything.

I've been working on Window Manager + other nice-to-haves for macOS 26. I do not need a model to one-shot the program for me. However, I am thrilled to get near instantaneous answers to questions I would generally have to churn through various links from Google/StackOverflow for.

Post reply on HN