411–418 of 418 posts
Earlier quoted context omitted.
Their ToS says differently. You can't argue with what's explicitly in their legal agreement.
> You can't argue with what's explicitly in their legal agreement. Sure you can, that's what courts are for
Earlier quoted context omitted.
Their ToS says differently. You can't argue with what's explicitly in their legal agreement.
> You can't argue with what's explicitly in their legal agreement Sure I can. I can even contest it in court (if I had the money). Some "legal agreements", TOS, etc. are even unenforceable and blatant abuses of the law. And what's more, I can even consider ALL such legal agreements bogus and demand that the law changes to now allow them.
Good luck trying to classify this one as such. There's no valid argument given the fact that users are attempting to gain access to an offer in a way that isn't applicable to them. It's tantamount to deception and stealing, going somewhere you were not invited as though you were and taking something that wasn't given to you.
Earlier quoted context omitted.
Why would they have that feature in claude code cli if it goes against the ToS? You can use Claude Code programatically. This is not the issue. The issue is that Anthropic wants to lock you in within their dev ecosystem (like Apple does). Simple as that.
allowed shell pipes doesn't necessarily mean they want loops running them. One of the economic tuning features of an LLM is to nudge the LLM into reaching conclusions and spending the tokens you want it to spend for the question. presumably everyone running a form of ralph loop against every single workload is a doomsday situation for LLM providers.
There's nothing in TOS that prevents you from running agentic loops.
This and threatening OpenClaw (now at OpenAI), Anthropic really on a roll making friends in Open Source. Previously discussed I think: Anthropic Explicitly Blocking OpenCode (173 points, 157 comments) https://news.ycombinator.com/item?id=46625918
Anthropic's position has been consistent from the start. Subscription plans are for first-party clients. Everyone else should be on per-token billing. That's really all there is to it. It has nothing to do with open or closed source.
Their documentation, cli, SDKs, TOS, their public stance - have major contradictions and are subject to so much interpretation that you can't really figure out what's allowed and what's not.
Opencode allowed? No
Custom agentic harnesses allowed? Yes
Opencode is a custom agentic harness? Yes, but not allowed
Which agentic harness is allowed? Can I use a custom agentic harness I develop - say, tmux that runs claude code? We don't know, but you'll know if we ban you
Earlier quoted context omitted.
Anthropic's position has been consistent from the start. Subscription plans are for first-party clients. Everyone else should be on per-token billing. That's really all there is to it. It has nothing to do with open or closed source.
Which means the plans are subsidized, and you pay with your data.
What's likely is that light users are paying for heavy users and API pricing is heavily inflated.
Earlier quoted context omitted.
> you've just reasserted "because they said so". Yes. That is all that matters. That is the contract you've signed with the business, end of story. They have their reasons (3P clients suck at caching, this is a subsidized tier to create subscription lock in, etc) but that's besides the point. If you sign a contract, you abide by its restrictions or you categorically abusing the offering. > There is another alternativ…
Ah, but the ToS I agreed to didn't state it.. they even admitted it wasn't clear and then unilaterally revised it. And yes, I agree businesses must abide by the ToS, it would be too much of a business risk not to.. but individual users, using for their own purpose can make their own choice. An individual choosing to not abide by their unfair ToS is perfectly reasonable and doesn't make them unserious or immature peop…
This is terrible from a product perspective. Right now they can subsidize workflows with poor caching performance.
If OSS hacked clients become a big thing, they can't do this anymore.
Earlier quoted context omitted.
Ah, but the ToS I agreed to didn't state it.. they even admitted it wasn't clear and then unilaterally revised it. And yes, I agree businesses must abide by the ToS, it would be too much of a business risk not to.. but individual users, using for their own purpose can make their own choice. An individual choosing to not abide by their unfair ToS is perfectly reasonable and doesn't make them unserious or immature peop…
> then the solution is to make it chew through the quota quicker This is terrible from a product perspective. Right now they can subsidize workflows with poor caching performance. If OSS hacked clients become a big thing, they can't do this anymore.
If Claude code can have great performance by optimised caching and Opencode doesn't.. I'll run out of quota using a third party client and have a terrible experience, and the choose to use Claude code instead because wit is superior.
So why not simple quota counting?
Earlier quoted context omitted.
> then the solution is to make it chew through the quota quicker This is terrible from a product perspective. Right now they can subsidize workflows with poor caching performance. If OSS hacked clients become a big thing, they can't do this anymore.
How does that make sense? If Claude code can have great performance by optimised caching and Opencode doesn't.. I'll run out of quota using a third party client and have a terrible experience, and the choose to use Claude code instead because wit is superior. So why not simple quota counting?
Consider this: you are Anthropic. There are some Claude Code used cases that will have poor caching performance. Let's say these are 10% of your use cases.
You explicitly don't count cache misses right now because it would make the UX poor for these use cases. It's no big deal since the remaining 90% of use cases can subsidize the 10%.
Now open source clients become a thing. Instead of 10% of usage having poor caching, it grows to 50%. You can no longer subsidize those users because the economics don't work.
You have to start counting cache misses and the UX goes to shit for everyone.