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

Anthropic downgraded cache TTL on March 6th

github.com

Post reply on HN
Print view View original

431–440 of 447 posts

Re: Anthropic downgraded cache TTL on March 6th

#431

Post by deaux »

oooyay wrote: post #301
deaux wrote: post #220

Earlier quoted context omitted.

I haven't seen a single case of that happening with Anthropic yet. Every time someone has gotten banned it's because they either used third party harnesses which went to great lengths to impersonate claude code (obvious evasion), or because they set things up so it maxxed out their usage 24/7. I'll change my mind when I see otherwise. And this isn't being positive about Anthropic support or their treatment of users,…

I plugged this question into Claude and told it to limit me to 10: 1. Cancer patient banned mid-payment — https://news.ycombinator.com/item?id=46675740 2. Hobbyist coder, VPN trigger, forms into void for 10+ months — https://news.ycombinator.com/item?id=47286867 3. "Reinstated" but still locked out — two systems out of sync — https://news.ycombinator.com/item?id=46007408 4. Banned for testing vision API — https://new…

Would've been nice if you'd read them as not a single one of them even mentions Agent SDK or claude -p usage, the topic of this thread.

Re: Anthropic downgraded cache TTL on March 6th

#432

Post by dwaite »

sshine wrote: post #112
rlpb wrote: post #80

Earlier quoted context omitted.

> It's all circumstantial but everything points towards "desperately trying to cut costs". I have been wondering if it's more geared at reducing resource usage, given that at the moment there's a known constraint on AI datacenter expansion capability. Perhaps they are struggling to meet demand?

> Perhaps Anthropic is struggling to meet demand? Yes, definitely, they’re gracefully failing to meet demand. They could also deny new customers, but it would probably be bad for business.

For my part, I've tried to help reduce their demand by cancelling my subscription.

Re: Anthropic downgraded cache TTL on March 6th

#434

Post by oooyay »

deaux wrote: post #431
oooyay wrote: post #301

Earlier quoted context omitted.

I plugged this question into Claude and told it to limit me to 10: 1. Cancer patient banned mid-payment — https://news.ycombinator.com/item?id=46675740 2. Hobbyist coder, VPN trigger, forms into void for 10+ months — https://news.ycombinator.com/item?id=47286867 3. "Reinstated" but still locked out — two systems out of sync — https://news.ycombinator.com/item?id=46007408 4. Banned for testing vision API — https://new…

Would've been nice if you'd read them as not a single one of them even mentions Agent SDK or claude -p usage, the topic of this thread.

I did read them but I interpreted the topic of this thread to be Anthropic's vague approach to compliance enforcement not specifically how claude -p is used and interpreted by Anthropic.

Re: Anthropic downgraded cache TTL on March 6th

#435

Post by zoogeny »

onoesworkacct wrote: post #416
zoogeny wrote: post #370

As an aside, I built a tool to manage my own chat interface over the provider APIs. I added caching because the savings are quite significant and I have a little countdown timer that shows me how much time remaining until the cache is expired. However, for the basic turn-based conversation the cache (at 5 minutes) is almost always insufficient. By the time I read the LLM response, consider my next question, write it…

give it a skill that runs a timer in the background and every 4.5 minutes says "ping? pong!"

Interesting idea. I suppose one could also have response settings (e.g. max response tokens) to ensure the model doesn't waffle on and run up costs. In a best-case scenario "ping" would be one or two input tokens and a "pong" response would be one or two output tokens, so the cost of the operation would be the preserved context size times the cache read cost (one could avoid doing a cache write since I believe the cache read would reset the platforms cache timer).

It would be interesting to graph the cost/savings of this approach based on context length, percent cached, etc.

The UI for this is a bit tricky, I could mark conversations as "active" and then do the ping/pong dance on only active conversations and up to some determined max cached (e.g. 1 hour).

Re: Anthropic downgraded cache TTL on March 6th

#436

Post by Dylan16807 »

infecto wrote: post #428
Dylan16807 wrote: post #404

Earlier quoted context omitted.

The lower tier openai and google plans don't have access to the same models. Where are you seeing popular plans that are simply token purchases?

I guess if you want to go that deep sure they sometimes offer early access, access to new agents/models but ultimately it’s a function of tokens. The selling point for most/all providers is x times the usage. You are upgrading for the token access. Claude was the topic at hand and higher tiers buy you more tokens. I know some like Gemini bundle a ton of junk alongside the tokens but you really are still buying yourse…

> $100 is the same as $200 except token count

But I was making an argument about the $10 plans, not the $100 plans.

Claude doesn't even go that low. Except the free plan which has a very reduced feature list.

Claude's $20 and $100 are pretty similar except tokens, that part is true. So they're a bit higher priced and more of the "it's just tokens" model. But the market as a whole is mostly selling a limited feature set down at lower price points. On average, getting up to the point where you have full access and are paying per-token is itself a premium jump.

Re: Anthropic downgraded cache TTL on March 6th

#437

Post by andriy_koval »

CamperBob2 wrote: post #352
andriy_koval wrote: post #206

Earlier quoted context omitted.

> Meanwhile the Minimax team just released a new open weights version of their 229B model yesterday. its under new license prohibiting any commercial use.

Who in the world would care about that, and why? You know what else was under a license prohibiting commercial use? The training data.

small shops are vulnerable against lawsuites, when will start spitting watermark text passages.

Re: Anthropic downgraded cache TTL on March 6th

#438

Post by mikenikles »

kfajdsl wrote: post #426
derwiki wrote: post #166

Earlier quoted context omitted.

Free VS Code is probably fine

These are like $20-50 subs, you’re probably paying your dev a hell of a lot more. Let them use the tools they want. I spend almost all of my time in Emacs or Cursor, but I still haven’t found a database client that I like better than Datagrip.

A database client better than Datagrip is a tough one, yet I'm attempting to do just that [1] :).

I'm in month 4 of development, working on it full-time.

[1] https://seaquel.app

Re: Anthropic downgraded cache TTL on March 6th

#439

Post by infecto »

Dylan16807 wrote: post #436
infecto wrote: post #428

Earlier quoted context omitted.

I guess if you want to go that deep sure they sometimes offer early access, access to new agents/models but ultimately it’s a function of tokens. The selling point for most/all providers is x times the usage. You are upgrading for the token access. Claude was the topic at hand and higher tiers buy you more tokens. I know some like Gemini bundle a ton of junk alongside the tokens but you really are still buying yourse…

> $100 is the same as $200 except token count But I was making an argument about the $10 plans, not the $100 plans. Claude doesn't even go that low. Except the free plan which has a very reduced feature list. Claude's $20 and $100 are pretty similar except tokens, that part is true. So they're a bit higher priced and more of the "it's just tokens" model. But the market as a whole is mostly selling a limited feature s…

You are standing on top of an ant hill and I still don’t fully understand your position. The original post was about the premium service Anthropic plans. There is no such thing, you are simply paying for more tokens. Hope that helps.

Re: Anthropic downgraded cache TTL on March 6th

#440

Post by Dylan16807 »

infecto wrote: post #439
Dylan16807 wrote: post #436

Earlier quoted context omitted.

> $100 is the same as $200 except token count But I was making an argument about the $10 plans, not the $100 plans. Claude doesn't even go that low. Except the free plan which has a very reduced feature list. Claude's $20 and $100 are pretty similar except tokens, that part is true. So they're a bit higher priced and more of the "it's just tokens" model. But the market as a whole is mostly selling a limited feature s…

You are standing on top of an ant hill and I still don’t fully understand your position. The original post was about the premium service Anthropic plans. There is no such thing, you are simply paying for more tokens. Hope that helps.

Any $200/month AI plan is premium. Hope that helps.
Post reply on HN