Since there's a lot of questions about what this means, let me explain.
Anthropic has two different products that are relevant here: the Claude API and Claude Code. The Claude API has usage based pricing. The more you use, the more you pay. With Claude Code, you can get a monthly subscription which gives you a fixed amount of usage. Comparing equivalent token generation between the Claude API and Claude Code, Claude Code with a subscription is much cheaper.
When it comes to third party products such as OpenClaw and OpenCode, Anthropic has made it clear those products should be using the Claude API and not the internal Claude Code APIs. OpenClaw and OpenCode have both been using the internal Claude Code APIs as when a user has a Claude Code subscription, the internal Claude Code API gives you tokens at a much cheaper rate than the Claude API. Presumably Anthropic makes Claude Code cheaper than the Claude API because they are willing to give users a discount for them to use Claude Code vs a competing product such as OpenCode.
It looks like until recently OpenCode tried to get around Anthropic's requirements by offering "plugins" in OpenCode that would allow users to use their Claude Code subscription in OpenCode. This PR mentions as much at[0][1]:
> There are plugins that allow you to use your Claude Pro/Max models with OpenCode. Anthropic explicitly prohibits this.
> Previous versions of OpenCode came bundled with these plugins but that is no longer the case as of 1.3.0
This PR seems to be in response to Anthropic threatening OpenCode with legal action if they keep using the internal Claude Code APIs.
I think we can attribute a bunch of consternation here to drift between assumed and actual licensing terms.
The actual licensing terms for Claude Code expressly prohibit use of the product outside of the Claude Code harness. If you want Opus outside of CC, the API is available for your use anytime.
Some percentage of the community seems to assume their Claude Code subscription licenses allow free usage of CC across any product surface - including competing products like OpenCode. While this is a great way to save on API costs, the assumption is incorrect. In fact, it is *so* incorrect that Anthropic has encoded their licensing terms into their Terms of Service, and a result can take legal action against any violating parties.
We can have separate discussions about Anthropic’s use of the Common Crawl in pre-training, or whether foundation labs adhere to robots.txt conventions. But those don’t directly impact Anthropic’s right to bring litigation.
——
Outside of that I think angry users have their own stated preferences v revealed preferences here. They claim they want Opus on their terms, and Anthropic’s actions infringe on their user rights.
Angry folks: Opus is right there! You just need an API key! The reality is you want Opus in your devtools of choice at discounted rates. You could at least be honest about your consternation
I think that’s a bit more nuanced. The actual „product” is not the harness, which is free anyway, but the Claude subscription. In any scenario, that’s what the customer continues to pay for. I understand why Anthropic is doing that, but I feel no need to defend it. Just like I understand why Apple limits your app choices to AppStore, but I’m not going to go out of my way to defend their decision.
It's way more nuanced, because the subscription is older then Claude Code - and they only started to have a problem with third parties using it after Claude Code. (And not with the release, just some time after the release)
>We can have separate discussions about Anthropic’s use of the Common Crawl in pre-training, or whether foundation labs adhere to robots.txt conventions. But those don’t directly impact Anthropic’s right to bring litigation.
Some of us don't care for Anthropic's "right to bring litigation" anymore than we care about some scumbag patent troll company doing things "within their legal rights".
We care for the morality of its conduct, the openess of its products, and the environment it creates.
I think this is disingenuous, people want to be able to use a tool that they pay for to do useful work on their own terms because they payed for it and don’t see the differential pricing model offered by Anthropic as legitimate.
> Anthropic has two different products that are relevant here: the Claude API and Claude Code.
No, the two relevant products are Claude API vs Claude subscription. There's no "Claude Code subscription". There's just a subscription for all Claude services at once.
The $20/mo Pro subscription only allows regular chat and Claude Code and does not allow you to export your API key without reverse engineering CC. The higher tiers allows console and direct API usage.
Basically, the concept of Claude-Code having its own API tier holds.
Sure you can. TOS docs are full of non legally enforceable wishful thinking bullshit, especially when they're written by an American company providing services to me in Europe. Most of the time they just expect (correctly) that they'll never get challenged in court over it.
Claude code might be subsidized but there are other risks
Like if any agent can use claude models then it exposes them to distillation risk. Where data gathered from millions of such agent usage can easily be used to train a model, making their model superiority subpar
Second thing is, to improve their own coding model, you need predictable input.
If input to their model is all over the place (using different harnesses adds additional entropy to data) then it's hard to improve the model along 1 axis.
Cache is money saver in computing. Their own client might be lot better at caches than any other agent so they do not want to lose money yet end up with disgrunted customer that claude isn't working as good
And also, if a user can simply switch model in an agent. Then what moat does anthropic have? Claude code will not include other companys models and thus will allow them to make their claude code more "complex" with time so the workflows are ingrained in users psyche to the point using anything else becomes very difficult and user quickly returns to claude code
They are not entitled to a moat, and their customers do not owe them one. Several companies have narrow or no moats. Dell and HP are two examples when it comes to their PC business.
This idea that companies should be allowed to lock down their products just so they can have moats, is how we ended up with printer ink being more expensive than crude oil or champagne.
Companies are absolutely allowed to lock down their own products. Netflix is a great example, you don't bring your own client for Netflix.
The whining/entitlement in this thread is ridiculous. The API is always there for you to use as you desire.
If you want to use the loss leader on the other hand, you agree to abide by certain terms. But if you don't want to do that, just use the API. It's not that hard.
> Cache is money saver in computing. Their own client might be lot better at caches than any other agent so they do not want to lose money yet end up with disgrunted customer that claude isn't working as good
I’d bet a reasonable amount that this could be the case. They are very well incentivized to maximize cache use when it’s basically not pay per token.
So they've been advancing in making the AI use the computer through the same API as a person (screen/cursor/keystrokes), and the dream is a Future where AI can use a PC and handle tasks and tools like a human user.
But to use their product,you have to go through the non-human-friendly API route, or else it's against the rules and Anthropic will sic their legal team onto you...
Something about this reasoning seems brittle. Specially in a world of Agentic tools
The part I never really understood, was I thought the subscriptions were to try and boost Opus usage, not claude code usage ? I'm not sure why they care whether you use API or claude, as they limit the number of tokens you can use anyway - and once the request hits the model, I would have thought it takes the same amount of effort to process it regardless of where it comes from ?
It’s definitely to encourage Claude code usage. Owning the interface through which your core product is delivered is a hedge against the commoditisation that everyone talks about. Eg, it’s much harder to switch from Claude code to cursor or vice versa than it is to switch between models in cursor (I sometimes don’t even notice model defaulting to composer inside cursor)
This is clearest reason for us to accustom ourselves to using open weight models on open source harnesses. Whatever advantages the frontier closed models offer, this will turn into ash in the mouth, when the enshittification cycle begins. And don't be mistaken, it will begin. There is no precedent which can claim otherwise.
I am sure the models themselves are being RLHF tuned to work very well with the proprietary agent harnesses. This is all turning into a huge trap right in front of our eyes and the target is not just programmers but also companies whose core product involves software production.
I can believe it - maybe they feel they have enough of a lead in usage with programmers with Opus that they want to locking down the tooling side as well.
I'm really struggling to understand how Anthropic is benefited by not allowing this. Its bad PR for no good reason. The only thing I can figure is that Claude Code is hemorrhaging money, they're too afraid to actually enforce reasonable token limits, and the only thing that's keeping it from totally bankrupting the company tomorrow is: controlling the harness and having the harness dynamically route toward Haiku or Sonnet over Opus when Opus is overloaded, without telling the user. Or maybe, they're extremely interested in observability of the exact prompts users are typing, and third party harnesses muck that data in with the rest of the context that gets sent, so its harder to detangle the prompt from the noise?
Like, in any event, I seriously get the feeling that Anthropic doesn't just not care about their users, but actively dislikes them. Like, they must be losing so much money on each Claude Code subscriber that if a million people all said "we're switching" they just wouldn't care. I get this vibe even from watching videos of people working on the Antrhopic team; like they all think they're Gods above mere mortals, serving some higher purpose, and nothing matters to them except Building the Machine God.
OpenCode is awesome. Claude Code is nothing special at all. Last month I switched to just using OpenCode with a Codex $200/mo subscription, and that's been great. Let the weirdos at Anthropic do what weirdos do, and hopefully one day their name is never mentioned again in polite society.
This is analogous to when Google launched Gmail with 1GB of storage and then a bunch of third-party apps cropped up that took advantage of it to use it as a generic online file storage drive.
There was GMailFS[0] and Gmail Drive[1] - this is before S3, dropbox, and a time where web hosting would give you ~10MB or so of space.
Google updated their ToS and shut down accounts using their service in ways they weren't intended via these apps - because obviously the 1GB of storage was a loss-leader into Google's ecosystem (and it worked)
Same thing today - "unauthorized" third parties taking advantage of a loss-leading[2] deal - complete with similar trademark violations to boot[3].
Google have more cash to burn in the AI race so can be more forgiving today in how their codex plans are used. Anthropic are still a private company and can't.
[2] it's a big q just how large a loss leader the max plans are considering a fixed harness, prompt caching etc. but point still stands. you're getting up to $5k of RRP tokens for $200
No, because in those cases you're still a user of gmail. When you tell people your email address, or send people email, and it contains "@gmail.com", you're still implicitly advertising for Google. From Google's perspective that's still worth the few KB per day of bandwidth and 1GB storage (which the vast majority of people never use the entirety of, anyway) they're giving away.
But when you use gmail accounts as file storage, you're both a higher-cost user and also doing nothing to further Google's ecosystem (since the email address itself is probably not being used for genuine messaging at all).
And here, you're still using Claude Opus, and when people ask you what you used, you'd say OpenCode client with Claude (Thunderbird client with Gmail).
there is nothing about claude code that prevents you from using it for non coding use cases. nothing that happens in open code or any harness for that matter is hidden from anthropic. neither does open code allow access in some nefarious use case that claude code does not.
the difference is not like the difference between gmail and gmailfs like you seem to be misunderstanding. a more accurate comparison would be the difference between curl, or httpie vs postman.
It's not analogous at all because Google intentionally provided interfaces for those clients and even instructions for using them.
An analogous situation would be if someone reverse engineered the Google Maps API and provided their own app that showed maps using the Google Maps data.
> And if Google Maps charged per tile viewed, so the user pays the same amount regardless of which maps client they used, would your opinion hold?
Yes. Why wouldn't it hold?
Anthropic has a pay-per-token API. You can use OpenCode with it.
Maybe my consistency comes from having worked with contracts and agreements in the real world, where the end user doesn't get to pick and choose which terms they want to abide by.
When you sign up to use a service, you're not signing up to use it however you would like, on your own terms. You're paying for a service that they offer. They are not obligated to continue offering it to you if you try to use it a different way.
My point is that model providers are just a compute service, and should have no say in what sends the data, or displays the data. Especially when they only bill based on the quantity of data.
They have an API for exactly that. You can use it.
They offer a separate plan with discounts for use with their tools. You can also choose to take advantage of those discounts with the monthly fee, within the domain where that applies. You cannot, however, expect to demand that discount to apply to anything you want.
You can argue about what you want it to be all day long, but when you go to the subscription page and choose what to purchase it's very clear what you're getting.
> They are basically a utility
Utilities like my electric company also have different plans for different uses. I cannot, for example, sign up for a residential plan and then try to connect it to my commercial business, even though I'm consuming power from them either way.
Utilities do not work like that. They do have contractual agreements about how you can use the resources provided.
This argument is predicated on Anthropic losing money on the subs, but I'm not sure that's a cut and dried argument. OpenAI have said publicly that they're (very) profitable on inference, and they're much cheaper than Anthropic. I suspect this is just artificially trying to create a moat. The problem is their moat is not as sticky as they think it is - I completely ditched Claude for Codex a while ago, my money now goes to OpenAI, and I'm very happy with it. For a while Claude was noticeably better, but that's not the case any more - in my case I prefer Codex.
They aren't public companies (yet). They are allowed to just lie about these things. It's also not really reasonable to only count inference compute as a cost since it's not like any of these companies could stop doing R&D without being abandoned for having worse models within a year or 2
Here's one possibility: Anthropic understands the value of the brand and the harness and that those two things are connected, specifically because they came from behind. OpenAI almost accidentally launched a global brand overnight. ChatGPT went from nothing to the kind of english word you hear in non-anglophone countries in about a month. Millions and millions have used it (at least once) and more people associate it with AI than use it. OpenAI's problem is managing the big industry links so that by the time the hype cools down, they're already plugged into tools. Their "moat" is that number of companies that matter is actually small and all those companies like predictable, enterprise shaped solutions with contracts and stuff. Unlike developers who might switch their subscriptions quickly and absorb the productivity cost of switching (or minimize that cost), these big companies don't want to be constantly optimizing compute vs rental rate. They want to convert an unruly value (programmer productivity) to something easy, not replace it with a scheduling or optimization problem.
That was working ok until Claude, specifically Claude Code showed up. This was a really useful code-writing harness (that also signed your commits, advertising itself to everyone) that took what are essentially very similar models and made Opus feel like the future of software while GPT 5.2 and friends are just code agents. The performance, ability to handle long term tasks, all of that was basically similar but the harness oriented the model to reason, shell out sub-agents, write scratch code, add console logs, all the sorts of things that 1. seem like science fiction, and 2. improve output a little. Then from fall of last year to no you don't have developers saying "look what I made with LLMs" or "Look what I made with AI" but "Look what I did with Claude". There are not very many blog posts out there about the future of software being re-written due to GPT 5.2 getting autocompaction, but that same feature spawned thousands of "oh shit!" posts in Claude.
That's not a more defensible moat than name recognition + small N for customers. It's a scarier position because if someone else figures out how to deliver the same result (Opus + sonnet + Haiku in a managed ensemble) in a way that was sharp and viral, the same thing they did to OpenAI could happen to them. They still supply the compute but the fact that anyone gives a shit about them is their harness makes it look like more and better code is being written. If that's your situation, you gently write the OpenClaw guy, you threaten to cut off and sue OpenCode for using subscription sign-in. You don't do those things because of a numerator/denominator problem with token cost and monthly fees. You do it because someone using your models in a better harness is a clear brand threat.
> Google have more cash to burn in the AI race so can be more forgiving today in how their codex plans are used.
Even despite the larger cash pile to burn, Google is in the middle of their own controversy around what many feel is a rug-pull around how Gemini "AI credits" work and are priced.
Theory 1: the internet has been fully strip mined for all content and is now dead. See that graph of StackOverflow questions dropping off a cliff to zero. Nothing much worthwhile is being added.
Theory 2: they are all unethical as fuck and definitely learning off your data. You would be insane not to - theory 1 means all your free training data is gone, but all that corporate data is fresh, ripe and covers many domains that the amateurs on the internet never filled. You have to launder it some way of course, but it's definitely happening.
Theory 3: winner takes all. I don't care for "Claude" and your wishy-washy ethics performance. ChudAI has a better model and harness? I'm gone this evening.
Having all the users, even if they are exploiting you for cheap compute with their own harness, is essential.
Good theory and insight. Seems like that’s setting us up for some epic big co vs ai co legal battles for covertly training off sensitive and internal big co data
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.
> allowed shell pipes doesn't necessarily mean they want loops running them.
insane that people apologize for this at all. we went from FOSS software being standard to a proprietary cli/tui using proprietary models behind a subscription. how quickly we give our freedom away.
I don't know why this is downvoted, see my nephew (?) comment [0] for a longer version, but this is not at all clear IMHO. I'm not sure if a "claude -p" on a cron is allowed or not with my subscription, if I run it on another server is it? Can I parse the output of claude (JSON) and have another "claude -p" instance work on the response? It's only a hop, skip, and a jump over to OpenClaw it seems, which is _not_ allowed. But at what point did we cross the line?
It feels like the only safe thing to do is use Claude Code, which, thankfully, I find tolerable, but unfortunate.
Or can you? It's my understanding that you cannot use your subscription with the Agent SDK, that's what the docs say:
> Unless previously approved, Anthropic does not allow third party developers to offer claude.ai login or rate limits for their products, including agents built on the Claude Agent SDK. Please use the API key authentication methods described in this document instead.
Though there was that tweet [0] a while back by someone from Anthropic that just muddied the water. It's frustrating because I feel like the line between the Agent SDK and `claude -p` is not that large but one can use the subscription and one can't... or we don't know, the docs seem unambiguous but the tweet confuses things and you can find many people online saying you can, or you can't.
I'd love to play around with the Agent SDK and try out some automations but it seems I can only do that if I'm willing to pay for tokens, even though I could use Claude Code to write the code "for" the Agent SDK, but not "run" the Agent SDK.
Where is the line? Agent SDK is not allowed with subscription, but if I write a harness around passing data to and parsing the JSON response from `claude -p '<Your Prompt>' --output-format json` would that be allowed? If I run it on a cron locally? I literally have no idea and, not wanting my account to be banned, I'm not interested in finding out. I wish they would clarify it.
You can easily automate OpenCode - more so than the basic Claude Code or Claude desktop app - in a way that automatically uses the maximum amount of subscription quota every cycle. And in an inefficient way that Anthropic can't cache on their end.
If you know anything about subscription models, you know that ALL of them are built on the fact that most of the users don't use the full capacity available all the time.
1. If alternate UX exists, the user can easily replace them with another model as soon as it comes out.
2. All 'all-you-can-eat' plans everywhere comes with clause. Whether it is lunch at a restaurant or it is token-proxy-providers who might think of reselling Max plan to individuals at 20% markup.
> Let the weirdos at Anthropic do what weirdos do, and hopefully one day their name is never mentioned again in polite society.
Woof, that is a bit harsh... :) OpenAI will also face the same problem. They are doing it right now because they need to stand out in some way.
I disagree that this path is inevitable for any service provider. Case in point: Google Drive offers substantially better per-terabyte pricing for storage than Google Cloud Storage; like 100x+ cheaper. Yet, Google Drive has a beautiful API that is used by everything from the Google Drive desktop client to rclone. This is how most apps work, and this frame of thinking about the internet has worked for 40 years because of course 95% of users are going to use the frontend the company makes for their backend; but that 5% of users left on the fringe are oftentimes the most valuable, they're the ones that are going to pay Google an extra $1000/mo for 50TB of storage, and as long as the internal unit economics are good to go, Google should want that. Less edge-cases they have to deal with on drive.google.com, more revenue, all good things.
I do fully expect the limits on these subscriptions to be brought down. But that's not the problem people have with Anthropic today, nor the problem we'd have with OpenAI when they have to eventually do it. That's just the way of things.
The problem is: These actions by Anthropic scream: "Our internal unit economics are going nuclear and we need to do anything we can to regain control."
Low-key: I think the DoW situation was an inflection point for their usage internally. It spiked up hard after that. Dario spent all of 2025 being told "you're not investing enough into compute", but really didn't listen because he wanted to be "responsible" or whatever, and now they're shopping to every provider trying to find compute and are being told that there isn't any.
I don't think it's any of that. It's plain as day to see that as the owners of the API they can see every request, collect every metric they need. The routing can easily be done in the API gateway layer and doesn't depend on the harness at all. Good harnesses are not performing acrobatics around user behaviour to improve anything. They just understand the provider API very well and translate tool calls really well. Then they get out of the way to let the magic happen.
This is about controlling user behaviour, resisting standardisation efforts, to keep subscribers hooked to their models. If they keep their users on claude code, they don't have to adapt to evolving harness standards when shown increased usage across different harnesses. They will try to keep everything on their terms, on their poor quality of engineering, and the users accustomed to claude behaviour will be left with no quick way to switch to other models.
This strategy will keep working as long as they own a top tier frontier model with enough marketing hype behind it to blind unsophisticated users.
My guess is that the telemetry data they can collect from interacting with claude code is the "secret sauce" behind a lot of the improvements we're seeing with coding models right now. Look at cursors Composer-2 release today. Clicking "accept" during plan mode, committing changes and pushing to a remote repo, etc. is a really strong reward signal.
Can't collect telemetry from applications you don't control.
Yup, agreed with this as well. Probably also why they've been investing so heavily in the desktop Claude Code experience; very hard to gather great telemetry from a terminal app.
Literally everyone is desperately trying to figure out why it's so bad and how to make it work consistently using harness etc. But in spite of this massive effort things always go awry after a while. Maybe in a year or two someone figures it out.
This is my theory. They don't want other harnesses to use this because it costs them more. I don't know exactly how OpenCode works, but I'm assuming when people are using this plugin they are mostly using Opus for everything while Claude Code really only uses Opus for writing the actual code. It uses Haiku and Sonnet for almost all of the tasks outside of writing code.
So it hard for them to control and understand the costs of subscriptions if people are using them on different hardnesses that do things that they have no control over.
Yeah, but that's just the model the main agent uses. The subagents aren't Opus. They are Haiku and Sonnet. Most of the token heavy work is offloaded to subagents because of this.
They're trying for the vertical integration monopoly.
The times it works, it works well for the company at great cost to society.
Imagine the world we'd have if comcast got to control your web browsing experience.
If ISPs got started today, they'd sell the open web at API prices that no one can afford. Then sell the ISP's lock-in 'internet' for a low monthly fee.
My question is why people who don't want comcast's internet think other vertical integrated lock-in is fine.
Our markets game only works for the benefit of society if we have fair markets.
VC-backed loss-leader dumping to starve competition model breaks the game.
There are a plethora of models that you can use with open code. Anthropic is well within its rights to limit third-party usage of services that violate the TOS. As a Claude code user I’d much rather have the very best experience on Claude code than the largest supportability matrix for Anthropic models.
as someone who has used codex/claude-code/opencode I can confidently say that "the very best experience on Claude" is not the one that Anthropic provides software for.
Well, the challenge for Anthropic's users is: While Anthropic has fantastic coding models right now, they have a bad harness in claude code and the claude desktop app. The best experience using Anthropic's models is OpenCode and Cursor. This wasn't true ~3 months ago, and may return to being untrue in ~3 months, that's how fast these things change, but right now this is the case. Unfortunately, Anthropic models in OpenCode/Cursor are tremendously expensive; and that gets at some of the leading theories on why CC has degraded recently; that Anthropic has been forced to dynamically route more of the agentic process onto Sonnet/Haiku, or reduce the Opus thinking budget.
For all of these reasons, currently, the meta is ChatGPT subscription on Codex or OpenCode. But, again, these things change every few weeks.
I don't think this is as clean-cut as just saying "Anthropic is in their rights" etc. Of course, they are, to whatever degree they are; the bigger problem is that you've got $100/mo and $200/mo Claude subscriptions who are actively saying "the $20/mo Codex subscription is better in every way", possibly because of these thinking budget/routing changes people suspect have happened this month. In other words: Anthropic is at-capacity after the DoW incident, they need to load shed, and they've chosen to harm their high-paying power users and Enterprise over just temporarily slowing growth a bit by hurting the $20/mo plan. And, frankly, they might be right: because for every $200/mo user that jumps over to ChatGPT, half of them will be back once Anthropic can scale capacity, and if they can gain 20x $20/mo users who only use half their sub, that's a win.
They want lock-in for their UI/X, presumably. If Photoshop ran in the cloud, I doubt Adobe would let you make an alternative front-end either. Not that I'm sympathetic to them.
Yes, this. They need as much lock-in as possible before IPO. Most likely less about cash flow and more about IPO story telling.
We'll know for sure when they add full OpenClaw-like features to Claude Code like remote channels & heartbeat support. Both are partially implemented already.
It's served from the cloud but it runs entirely on your PC (except the AI generative tools). It can't run entirely offline though, because the js, webassembly, and assets are served chunked as-needed.
Saying Photopea is good enough is really underselling it. It's so far ahead of anything offered by the open-source community.
I'd be so happy to pay for a fully offline version of Photopea!
> I'm really struggling to understand how Anthropic is benefited by not allowing this. Its bad PR for no good reason. The only thing I can figure is that Claude Code is hemorrhaging money, they're too afraid to actually enforce reasonable token limits, and the only thing that's keeping it from totally bankrupting the company tomorrow is: controlling the harness and having the harness dynamically route toward Haiku or Sonnet over Opus when Opus is overloaded, without telling the user.
I have noticed odd behavior when choosing a model, it automatically switches after sometime. Some tasks do not require lots of power, so when I select Haiku, few prompts later, I see Sonnet popping up in the cost/spending. Happened few times.
Agreed. What I suspect is: the dynamic model routing on CC is way stronger than people realize, and that "Percent-based usage" is intentionally vague because while it is probably measuring "200M tokens per week" or something, they don't want you asking questions about whether you're getting 200M Haiku tokens or 200M Opus tokens. A token is a token to the usage limit, where it comes from doesn't matter to the usage limit. But, to OpenCode it might, because OpenCode can just fire-and-forget everything at Opus (and probably does).
> OpenCode is awesome. Claude Code is nothing special at all.
It's weird, my experience is the exact opposite: OpenCode had the weirdest issue of interpreting C:\Users\some-user\Documents\Projects\... as a bunch of escaped characters without \ and created some trash files even with Anthropic models. Before then, OpenCode had the issue of me copying multiple lines of text into it from clipboard somehow quitting the program (crashing?) and then making every following line (of text, notes etc.) be passed to the shell directly. Happened in both Terminal app and also Git Bash on Windows. It might just be that Windows is a shit OS, but Claude Code doesn't really seem to exhibit those problems (and neither does Codex, in their defense). I hope OpenCode has fixed at least some of those since.
> Let the weirdos at Anthropic do what weirdos do, and hopefully one day their name is never mentioned again in polite society.
What an odd thing to say. Maybe I'm out of the loop about some drama (aside from Anthropic having a bit of a spine in regards to govt. while being otherwise flawed still), but pretty happy with my Max subscription, they even fixed the Code GUI mode in their Electron app recently, it finally feels usable and I don't just need to juggle 4 terminal sessions.
One possible factor is reinforcement training: by forcing users to use their own CLI, they gather usage and training data which serves as feedback training data for their own tools. Users operating with a different agent with different tool implementation and quirks might poison that data.
The sibling reply given by strictnein is very likely a factor.
You are wayy overestimating the negative press. HN commenters are a negligible fraction of Anthropic's user base, and I can guarantee that even people here will forget about this and go right back to using Claude Code in a couple days when there is something else to be outraged about. The company needs to do what is best for its business.
The business driver I assume is not lock-in on UX (as some say here) but the additional signal Anthropic gets when using their harness vs a 3P one. It makes sense to discount the price if that signal helps you improve your models, but that discount makes no sense if the user is running your model in another harness and you just get regular API usage signal.
Since I discovered pi I cancelled my Claude subscription and subscribed to ChatGPT. On one hand, the competition is making miracles. On the other hand, it's pretty dooming that there is only one (1) company that keeps my agent cost reasonable.
Extensible coding agent written in typescript. It’s exactly what you (I’m projecting) want out of Claude Code if you’re okay investing time into building your harness or prompting an agent to build it.
> they must be losing so much money on each Claude Code subscriber that if a million people all said "we're switching" they just wouldn't care.
You're looking at it completely wrong. Claude Code is Anthropic's flagship product, not the API. They want to attract as many users as possible to Claude Code and lock them into their ecosystem, so they can squeeze them later. All of their questionable actions surrounding Claude Code and its subscription are ultimately in service of this goal.
The subscription isn't some kind of charity, it exists specifically because they know the average user isn't willing to pay the exorbitant API prices to vibe code their groundbreaking new B2B SaaS idea, but they want to capture that market share anyway, because it's the core of their long-term strategy. The subscription arose from that: it's a form of predatory pricing designed to attract as many users as possible while they still have VC money to burn.
Once that runs out and the time comes to IPO and start making real profits, they are going to increase the price drastically, and what's where the lock-in comes into play. If everyone is using some open-source alternative that natively supports every other provider on earth, they will be far less likely to continue paying for Claude specifically instead of just switching to a competitor. Not to mention, they'd also lose out on the free advertising from things like CLAUDE.md and the commit co-signing (because that's all those things are, the only reason Claude Code doesn't support AGENTS.md is because CLAUDE.md serves as an advertisement in public repositories).
> like they all think they're Gods above mere mortals, serving some higher purpose, and nothing matters to them except Building the Machine God.
This is all just part of their marketing strategy, and you shouldn't read too much into it.
Anthropic doesn't care and it is all to look good for their IPO.
They are still losing billions of dollars and will do anything to keep people hooked onto the API and will litigate against their own customers.
They will even lobby against open-weight models which is their biggest threat and want to make them illegal to run in the US just for them to succeed.
Anthropic are not your friends and want you to become addicted / over-reliant on Claude Code (hence the free $20 spins at the roulette until March, 27 2026) and charging others on their overpriced API.
Yes, it is true that companies often litigate against customers who violate their Terms of Service. The TOS is put into place to protect the company’s interests from user abuse.
Paying customers of Claude Code don’t receive a free-use license for any desired application. They’re paying to use Claude Code. Anthropic can take steps to litigate usage outside of those terms, even if customers find that fact really annoying.
This seems harsh and unfair. You aren’t allowed to stream Netflix through a third party streaming service because then the experience is controlled by the third party and there is no lock in to using Netflix and thus no benefit to the subsidy they give you in the plan.
OpenAI is allowing it as a PR stunt and because they have seemingly unlimited cash they can throw at user growth.
Anthropic sells a service that bundles server and client. They are not wild about people taking their server part and using their own clients because the business model relies on both client behaviors (Claude code does a lot of work to achieve > 95% cache hit rates; third party clients likely don’t) and flywheel of usage data.
If Microsoft went after third party clients that emulated M365 and used their backend, would that also “make friends” in open source somehow?
I mean, Microsoft started out by 'not making friends' in open source at the start. Then migrated to the embrace, extend, extinguish days. The current Microsoft is rather friendly towards open source compared to those days.
Typically services that try to gatekeep standard http don't get very far with the kind of people that like to modify their software.
Anthropic stance has not been consistent. There are even now still articles by them telling users about unattended use like in CI which would now be considered TOS
Their stance has just been clear for a year now, even if they've changed their mind in comparison to the introduction of the sub, which significantly predates Claude code
Or that API prices are inflated. We don’t get to see what their internal financials look like. My guess is that your guess is more correct but it is unclear what is actually happening.
Dario has stated multiple times he doesn't believe there is any value in open-weight models. Not surprised. This is not the behavior of an innovative company - it is fear-driven. They are seeing a rapidly shrinking moat.
This exactly. Kimi 2.5 has coding performance hardly discernible from Claude. The only way to maintain a business edge is to crush open source clients to force people into a closed ecosystem. Once there, create context moat where people are not in control of their own context data (cannot export it to open tooling). Maybe we can call it the Oracle play?
It’ll be interesting to see if companies get tricked. I think it’s inevitable that it goes like MySQL/Postgres, where the open tools gets way better
This is, I'm sorry to say, simply not true. Anthropic and Open AI are materially ahead of every open source model out there at this time. The best they can hope to do is be Sonnet-adjacent, and even then I have not seen it.
> are materially ahead of every open source model out there at this time
They aren't. Any difference is in sampling parameters and post-training flavor choices. These aren't things that are "materially ahead", that's basically just LLM themes.
Listen, I want more open weight models in the world. They create entrepreneurial opportunities and support use cases which the foundation labs don’t want to support.
But open weight models are consistently three to six months behind on performance compared to closed models, as confirmed by both benchmarks and personal use. They’re closer on coding and much further away on non-coding tasks.
There are theories as to why these models lag, which I won’t get into. But anyone claiming open-weight models are close to closed-weight models is ignoring significant evidence to the contrary.
What's amazing is that LLM technologies are so immature that even basic engineering diligence isn't being done. (Like detecting token loops, for example.)
The onus isn’t on me. It’s on anyone contradicting findings by most benchmarks, because most of them show a clear advantage for Opus and GPT over OSS models.
I agree about Kimi 2.5. Also, MiniMax M2.7 that just dropped is amazing, and it is just a 200G MOE model and inference is very fast. I tried using MiniMax M2.7 twice today as the backend for Claude Code and it did very well for both existing Python and Common Lisp projects. I will try MiniMax M2.7 next as the backend for OpenCode.
If you believe benchmarks, maybe this is true. But I've done my own experiments and it is absolutely not the case for real world usage. The quality of output from Claude (Sonnet) was much higher than Kimi K2.5.
I agree. Kimi K2.5 is basically a clone. But also GPT 5.4 has replaced Opus for me. I have unlimited access to all of the proprietary models so I really only make the choice based on convenience.
thats only because kimi 2.5 was trained using data stolen from claude. it wouldnt exist without riding claudes coat tails. none of the so called 'open source' models would
That's not true, some open weight models didn't distill Claude or other then frontier models. E.g Llama. Yet achieved comparable performance (back then in llama's case).
If distillation wasn't a thing, they would certainly exist, they would have trained them from scratch or via a decent base models to remain economically viable.
What's for sure is that Claude wouldn't exist if it wasn't for data stolen from millions of creators. As they found themselves admittedly guilty of.
> E.g Llama. Yet achieved comparable performance (back then in llama's case).
At not a single point in history did Llama ever achieve comparable real-world performance to frontier models. I was around. At best they were earlier at benchmaxxing than the others.
Boo hoo. Claude was trained using data stolen from the collective works of all of humanity. If someone does it faster and cheaper by skimming the cream off the top of Claude then surely that’s just a market efficiency in the thieves business?
The dude built a mass plagiarism machine and wants to now profit off of his mass plagiarism machine, of course he's going to have antidemocratic ideas regarding people + technology.
His engineers told him "I don't write code anymore. Claude writes the code, I edit it, I code around it".
In 6 months, people won't work anymore. They will all use my products, outsource the thinking, why bother.
Oh and open weight models have no value...
There is a paper out there showing 30% of CEOs/C-suite have some psychopath tendencies. Not sure if they even used the term narcissistic , but I would add delusional.
IIRC it was called Clawdbot when Anthropic complained. IANAL but I believe the holder of a trademark is obligated to defend it against infringement. Hard to say that Clawdbot was not potentially infringing, given its purpose. It's not clear how much leeway Anthropic had given his initial choice of name.
I still think Anthropic should've bought Clawdbot/OpenClaw. Feels like a missed business opportunity to expand your market share by capitalizing on the hype.
"now at OpenAI" were my original words - they did the equivalent of an acqui-hire and "protected" OpenClaw in a foundation.
In the context of the seemingly aggressive machinations of Anthropic your hair-splitting without clarifying beyond "OpenAI didn't buy OpenClaw" seems itself misleading and rather counter to helping conversations progress.
It could reasonably cause confusion in the marketplace. Again, IANAL.
If you wanted to find out the actual legal arguments, you could release and promote software for white-collar workers called Mycrosoft Offyce and I am sure you will get official legal answers from Microsoft's counsel.
The example in question is a Trademark for "T. Markey" and the conflicting mark is listed as "Tee Marqee".
Reasonableness is a thing here. Every person in this thread knows why it was called Clawdbot and not (say) Peterbot or cronbot. That we all know that reason is the problem.
Edit, the USPTO does in fact make this blindingly easy to find:
"How do I know whether I'm infringing?" --> "The key factors considered in most cases are the degree of similarity between the marks at issue and whether the parties' goods and/or services are sufficiently related that consumers are likely to assume (mistakenly) that they come from a common source."
The OpenCode guys have really surprised me in the way they've reacted to Anthropic shutting down the side-loaded auth scheme. Very petty and bitter. It's clearly just a business decision from Anthropic and a rational one at that, usage subsidization to keep people on the first party product surface is practically the oldest business move in the book and is completely valid.
Ever since the shutdown of the side-load they've been pretty vocally anti-anthropic on twitter. Paranoid that anthropic is going to torpedo them via some backdoor now that they own bun, insinuating that anthropic shut down the auth from a position of weakness since OpenCode is a superior product, etc.
The thing is OpenCode IS a great product, I'm not sure it's "superior", but unfortunately the way things are evolving where the model + harness pairing is so important, it does seem like they are in a similar position to Cursor (and do not have the resources to try to pivot into developing their own foundational model).
I wouldn't call OpenCode a "great" product tbh. It's nice that it's FLOSS of course, but the overall quality is a bit underwhelming and it's clearly possible to build much better open agentic harnesses. It would be nice if more people tried to do this.
I think frankly OpenCode is delusional to think that Anthropic is actually "concerned" with them in any way. Anthropic's concerns at this point are on the geopolitical level. I doubt stamping out ToS-violating usage of their subscription services is even on executive radar. OpenAI only allows it because it's a cheap PR win and they take those where they can get them.
Yeah, I recognized the PR author from Twitter (same avatar) and man he really does come across as incredibly juvenile. Shamelessly talking up OpenAI while shitting on Claude models and the motivation is just so transparent.
This has been explained many times in this thread. Your subscription to Claude models for use in Claude Code is subsidized. That is, it is only meant to be used with that harness.
When you use that API key with OpenCode, you're circumventing that.
The PS5 is subsidized because the make money with the games.
Printers are subsidized because the make money with the ink.
The API use is subsidized because they make money with Claude Code?
I would understand if Claude Code could only be used with Anthropics API but not the other way around. 1 million tokens is 1 million tokens unless Claude Code is burning tokens and others are more efficient in token use.
I'd say that they want Claude Code to become the standard, so that they can milk corporations on enterprise plans. We individual subscribers are nothing, but we'll go to work and be vocal about specifically having Claude.
Valid question. It's because they have a separate product intended for use with general tools: Their API.
Their subscription plans aren't actually "Claude Code Plans". They're subscription plans for their tool suite, which includes claude code. It's offered at a discount because they know the usage of this customer base.
OpenCode used a private API to imitate Claude Code and connect as if it was an Anthropic product, bypassing the need to pay for the API that was for this purpose.
Anthropic has been consistent on this from the start. The subscription plans were never for general use with other tools. They looked the other way for a while but OpenCode was openly flaunting it, so they started doing detection and blocking.
OpenCode and maintainers have gone on the offense on Twitter with some rather juvenile behavior and now they're trying to cheekily allow a plugin system so they can claim they're not supporting it while very obviously putting work into supporting it.
Most of the anger in this thread comes from people who want their monthly subscription to be usable as a cheaper version of the public API, even though it was never sold as that.
Because models are quickly moving toward commoditization, whether the big three like it or not. The differentiator now is tooling around those models. By eliminating OpenCode's auth stuff, they prevent leaking customers onto another platform that allows model choice (they will likely lose paying customers to one of the major inference catalogs like OpenRouter once they move from Claude Code to OpenCode).
Why does Netflix care how the movies they stream to you are consumed? Shouldn't your $8/mo allow you to stream any movie to OpenFlix and consume however you like?
You are also not allowed to show these Netflix movies on a big screen in front of your house and charge people. The 8 dollar are for a specific use case, just like the tokens in the subscription.
If you use Claude through an interface that’s not Claude Code, you’ll only stick with it for as long as it proves itself the best. With other interfaces, you can experiment with multiple models and switch from one to another for different tasks or different periods of time.
Those tokens going to other providers are tokens not going to Anthropic, so they want to lock you in with Claude Code. And it clearly works, since a lot of people swear by it.
because he is giving them at 90% discount in their subscription.
they are more than happy if you use the tokens at api pricing, but when subsidized they want you to use their claude code surface.
Source: i run pretty much all of these agents (codex, cc, droid, opencode, amp, etc) side-by-side in agentastic.dev and opencode had basically 0 win-rate over other agents.
Anthropic provides subsidized access to Claude models through Claude Code. It is well understood to be 'a loss leader' so that they can incentivize people to use Claude Code.
OpenCode lets people take the Claude-Code-only-API-Key, and lets them use it in a different harness. Anthropic's preferred way for such interaction is getting a different, Claude API key (and not Claude Code SDK API key).
---
A rough analogy might be something like getting subsidized drinks from a cafe, provided you sit there a eat food. What if someone says, go to the cafe and get free drink and come sit over at our cafe and order food. It is a loose analogy, but you get the idea.
If it wasn't the case, the Claude API pricing would be the same, $200 for unlimited use. But it's metered.
We don't know if Claude Code bleeds money for every user that touches it. Probably not. But the different pricing is a strong enough clue that it's an appeal product with subsidized tokens consumption.
API is intended for a different audience - companies with a big pocket who aren't as price sensitive as private users. So the pricing will be different than for a private subscription.
There is huge value in getting people to subscribe to recurring payments. Giving people a discount to do so makes sense and does not mean that the subscription service loses money.
Is this what the legal request demanded or is this just something that OpenCode is doing out of spite? Seems unclear. To me the meat of this change is that they're removing support for `opencode-anthropic-auth` and the prompt text that allows OpenCode to mimic Claude Code behavior. They have been skirting the intent of the original C&D for awhile now with these auth plugins and prompt text.
Using your API key in third-party harnesses has always been allowed. They just don't like using the subsidized subscription plan outside of first-party harnesses. So this seems to be out of spite
"Legal action" means you filed a lawsuit. This looks more like someone sent a list of requested changes, backed up by an implicit or explicit threat of legal action.
That's how these things usually (should) go; a good legal system (be it civil suits, insurance claims, mediation, etc) will only actually take on cases if you've tried a reasonable approach first, e.g. asking nicely.
Fun fact: In Germany, the civil courts will usually take the case anyways if it has merit, but the winner ends up paying for the whole lawsuit if they failed to make an effort to resolve the case before suing.
Worth noting something the PR comments are glossing over is that the auth plugin wasn’t actually deleted in this PR, just removed from the default bundle. A third-party plugin doing the same thing is still technically on the table. OpenCode just can’t be the one shipping it anymore. Whether that holds up legally is anyone’s guess, but it’s not a hard wall.
I believe the solution to this kind of problems is that enough legislators (thinking of USA, EU) make it illegal to require a certain implementation of a software client to use an online service, whether this restriction is built into the communications protocol or checked through a side channel. Then the owner of the client device decides which implementation they'd like to use - so corporations can keep requiring the proprietary "official" one on their own corporate devices, but customers / end users who own their devices get to choose. Like, basic rights that come with ownership.
This would essentially take us back to what online services were 20 years ago, by outlawing a business model that relies on providers controlling stuff on users' devices. What's on the server is the company's business, what's on the client is the user's, and the boundary is clear.
On the one hand, services that would persist would likely no longer have free tiers (which would essentially mean free lunch for customers); since it's a commercial service run by a company that has costs, it's just normal to pay for it as a user, there's no free beer anywhere.
On the other hand, by paying for something, you'd get that thing and nothing else, as there's always going to be a client that doesn't leak data not necessary for the service to perform whatever you paid for, or impose arbitrary restrictions on its use. If any, these need to be server-side.
This just increases the price for everyone. The API is always there if you don't want to use the loss leader. The entitlement in this thread is insane.
This might be a very dumb question. But does this restriction also include the usage of claude code print mode with the -p flag and —continue or —resume?
This way I can use CC inside any application? Or what exactly was OpenCode using?
Businesses exercise power and control in the market. The purpose of this is to set a precedent (perceived or actual) — the auth system was a product, not an API. Anthropic is drawing the line between 'built on us' and 'built around us.'
I don't necessarily see this as an evil action. It doesn't inhibit open source, it sets terms of service and practice boundaries.
Granted this is a wildly unpopular approach, worse has happened in the OSS world...
Anyone know why OpenCode is integrating to ClaudeCode in the manner they were? Ie CC gives you an SDK, and i get the impression that Anthropic is fine with you using whatever external tools you want with the SDK .. otherwise why'd they publish an SDK?
So if CC has an SDK, why doesn't OC just use the SDK? I assume there's some functional reason why it doesn't perform to their needs? Maybe it's not low level enough? I'm unfamiliar with what sort of functionality a harness needs.
It makes me nervous as i'm using the CC SDK for my own wrapper though. Hypothetically what i'm doing is no different than embedding CC into an IDE.. though. Fingers crossed.
Users will say this-or-that about choice etc etc. It’s about subsidized tokens. Otherwise th users (and OpenCode) would have stopped pushing the workarounds months ago.
Yes, i completely agree - which was my question. To my understanding OC was using CC API directly instead of using the SDK. My question is, why? Does the SDK prevent some functionality they needed?
This is really sucks, it’s a bad move for Anthropic, OpenAI is allowing using their models in third parties apps and the china models as well …
I don’t believe Anthropic will win this battle.
I just want to use the tools in a way and customized way that I want, Anthropic not allowing me to me to do that forced me to use the codex models, thanks for that I’m very happy with the results. I’m cheering for OpenAI
OpenCode is an agentic coding CLI, similar to Claude Code, Codex etc. It supports different LLM providers, including Claude.
It originally had support for copying over your Claude subscription token and calling the same backend APIs as Claude Code, which meant you could use OpenCode with your Claude Pro/Max plans, but Anthropic came out and said that the plans were only meant for first-party clients, and everyone else should be using per-token billing.
Now OpenCode is removing this option from their product, it seems because of legal threats from Anthropic.
OpenCode is a agent .. harnest? The CLI or UI tool you use as an AI agent, similar to ClaudeCode.
You where not allowed to use your ClaudeCode subscription with other tool then ClaudeCode. I'm not sure if this is what got removed or if there is more too it.
I can understand how and why they are able to go after users, who violate ToS, but I do not understand how are why they can go after open source developers, who just publish some source code that seems to be protected speech under the First Amendment [0] in the US jurisdiction.
As much as it kills the living shit out of me to say this... Apple M5 Ultra will probably mean for a lot of us that we wont need openai or anthropic for coding llms anymore
While the M5 is impressive, its capacity to compete against enterprise models like Opus, Sonnet etc is off by a couple orders of magnitude. Even the most sophisticated open source models top out at 100-200 billion parameters whereas Claude, OpenAI are north of a trillion parameters.
While SOC is definitely the future or local AI, even if you get an M5 that's jacked to the tits, you still won't be able to store the entire model in unified memory on top of the OS and whatever other applications you have running. 128GB is the upper limit for unified memory on the M5, which on paper could support a model like gpt-oss:120b but still with a nerfed context size and quantised at that. Furthermore, the cost of a maxed out Macbook Pro M5 Max is between $8-10k depending on your storage option, screen size etc., so we can safely assume that the M5 Ultra will be even more. There's also no guarantee that the Ultra will offer double the amount of unified memory, it may only offer more cores but cores aren't the current bottleneck for local AI, memory is.
If you consider what you'd be paying above and beyond your requirements barring local AI, it would be adding $5-6k to the price tag at a minimum. That equates to 5 years worth of a Claude Code subscription! Even if you shouldered that cost with your NFT fortunes you likely wouldn't achieve performance parity with CC.
I'm equally as excited as you are about the future for local AI and I am actively working in this space everyday to improve it but we're still a long way off from being able to match model size, context size, token/sec, TTFB etc. A single H100 is so OP, so hosting thousands in a data centre it's expected that it should remain unrivalled.
The area where I'm having success with local AI is by pairing local models with other supportive technologies like databases and the like to compensate for smaller context size. There are still many in roads to be made in this area and that bodes well for the future of local AI as models become more efficient and sophisticated.
Anthropic subsidizes the fees when devs use their API within their own Claude code product. They do not give any such discount when the devs use their API outside their product. Seems fair enough. But looks like lots of developers expect the subsidy even when used outside their core product. Not fair, I guess.
In the end, a business gets to decide how and when to strategically subsidize their users.
Something that usually gets missed in these discussions is that the subscription quotas seem to rely heavily on prompt caching to be economically viable, or at least less unviable. They can and do have permutations of the system prompt, tools, skills, etc. that makes the first 20k or so tokens hit the cache and not use inference resources for that portion. In addition, from my monitoring, Claude Code with Max has about an 80% cost reduction via caching (equivalent if you had done the same work with API billing), and has been improving over time. If cache use passes on a discount of 90% I think it's fair to assume the actual cost to them is close to negligible.
So they're being obtuse about it for some reason, but if you want an economically sustainable model for AI companies they have to have some kind of optimization for the otherwise ridiculously discounted subscriptions. They sell subscriptions at the same rate and quotas to enterprise now, minus the $200 tier, so this isn't just consumer marketing being subsidized by b2b revenue.
Whether they're making money or just losing less, you can only get those kind of cache optimizations when you have a fixed client.
Maybe they could charge API users less if they use the same prefix that Claude Code uses? More coding agents using the same prefix results in better caching, reducing costs.
It's not clear what exactly the "legal action" is based on this github link. My pure speculation is Anthropic's lawyers have come up with a liability story boiling down to OpenCode helping end users violate the Anthropic ToS (i.e. tortious interference with contract).
Doesn't even need to be threatening, a notice of "this thing you're doing is in violation of our terms of service" should be enough... although I suppose that can be construed as threatening already.
Depends if it's existential, like I said. If my whole company depends on X and replacing it is intractable, there's not much other choice. Having looked at the landing page though, seems like they can just go with other models and it will (largely) be fine, yes.
teams enterprise customer here, we got knocked out too. the most irritating thing here? OpenCode is necessary given how incredibly buggy Claude Code is. I know Boris is a "hot shot" but the whole individual contributor developers running rough shot on that codebase + they still haven't really fixed the cache bug makes this a really poorly timed move on their part. I can't prove this, but I would bet management doesn't actually use Claude outside of the Desktop surface. They'd have known how much this would break if they were really using it at scale.
A few months ago, I had my Anthropic Claude Max account nuked for using OpenCode. That sucked, but I just opened a new Claude Max account under a different email. Which, yes, after re-reading their terms is also forbidden. But I had been playing by their rules ever since. Only using Claude Code and their official apps. So they got what they wanted - compliance.
Today they nuked my account again. I can only assume it was because I had the gall to find so much value in their product that even after they banned me once, I still wanted to give them money!
I've been around this planet a long time and I have never encountered a tech company as hostile to their users as Anthropic. And that includes Microsoft back in the 90's & 00's.
I really hope they change their ways. But for now, I'm done with them. I'll take my business elsewhere.
And the downfall of anthropic starts, OpenAI has had this all in the bag the whole time. Anthropic is a poor imitation of Sam's Master plan, it was over before it even started. Money grubbers, the lot of em!
Don't give them ideas please. They'll ask for more investment to do exactly this.
I miss the days when open source was a way to get your product in the developers hands and build trust. Stuff like this shows that the tide has shifted to primary focus on shareholders and potential hold on patents and trademarks.
Me too. I also miss the days when I was proud of my little open source projects. Now I just regret donating fuel, even a miniscule amount in the grand scheme of things, to the soulless lawnmower that has already chopped down so much of my joy in work and promises to eventually shred the paycheck, too.
I hear yah, especially knowing that AI crawlers just don't respect ROBOTS.txt or anything similar, but there's still nothing wrong with writing code for fun.. No need to lose that!
avoid bun is my take away... if anthropic decides you're a competitor and with the way AI is evolving you will be a competitor soon - don't rely on any anthropic tools or models.
Anthropic is a shit company. I cancelled my subscription 2 years ago once they started calling for regulation. They might have gotten folks to side with them in the OpenAI debate, but they are just another shit company like OpenAI.
The people mad about this feel they are entitled to the heavily subsidized usage in any context they want, not in the context explicitly allowed by the subsidizer.
It's kind of like a new restaurant started handing out coupons for "90% off", wanting to attract diners to the restaurant, customers started coming in and ordering bulk meals then immediately packaging them in tupperware containers and taking it home (violating the spirit of the arrangement, even if not the letter of the arrangement), so the restaurant changed the terms on the discount to say "limited to in-store consumption only, not eligible for take-home meals", and instead of still being grateful that they're getting food for 90% off, the cheapskate customers are getting angry that they're no longer allowed to exploit the massive subsidy however they want.
Anthropic has every right to place rules around their generous subsidization of the Claude subscription plans, which give limits of ~8-12x as many tokens as you'd get for the same expenditure in the PAYG API.
That said, demanding an open source repo remove information that Anthropic openly publishes and distributes for free (the prompt) is a bit odd...
This argument has been decapitated countless times already on HN. Anthropic already enforce usage limits for everyone. If those limits are higher than what they want users to actually consume, that's Anthropic's problem.
This move is anti-competitive and Anthropic knows it. They're hurriedly trying to lock the gates and lay landmines behind everyone after a massive surge of new subscribers so that they're stuck using Claude Code. They see it as vital to their survival to not just to be the gas pump for tokens, they need to control the platform.
I'm baffled how people don't seem intellectually able to grasp what you described here. Claude Code users on Anthropic subscriptions aren't subsidizing those using other harnesses because usage limits aren't counted on the harness layer. It's an anti-competitive move against vc-backed commercial harnesses like Opencode (vc-backed) or Openclaw (openai-affiliated).
> Anthropic already enforce usage limits for everyone. If those limits are higher than what they want users to actually consume, that's Anthropic's problem.
I mean, OpenCode is the one changing their app here. So it kinda seems like it's actually everyone else's problem.
> This argument has been decapitated countless times already on HN.
No it hasn't, because the argument is completely correct, and the people mad about it are mad they can't have unlimited usage instead of paying the token API prices.
> This move is anti-competitive and Anthropic knows it.
No it isn't, that's not what "anti-competitive" means, and no court in the world would label it as such. You can't go flailing around looking for legal jargon to attach to behavior just because you don't like it.
API is intended for massive scaled operations (companies) and has no hard usage limits, a subscription is intended only for individual usage (solo dev) and has therefore hard usage limits. Is it that difficult to grasp the difference between API and subscription models?
I must be alone in this but I don't think its heavily subsidized. I see their models as really overpriced. No way they cost that much. Could they really?
Cost to the business and price to customers are not the same thing. Even if it cost Anthropic nothing to run any of these (it actually costs quite a bit in electricity, infrastructure, ops teams to keep everything running smoothly, and above all else, extraordinary R&D expenses to develop the models), they could set the price at a million dollars per token if they wanted to.
That clarified: yes, every major lab is losing money on full utilization of their inference subscription plans. The API prices are what the business has determined they need to achieve profitability, and are not reflective of actual costs as you point out, but the discounts vs API pricing can get pretty extreme. Some users report 50x+ (98%+) discounts on the $100/mo Max subscription plans vs PAYG API pricing¹. Even the skeptical, contrarian takes that focus on cost to the business will tell you that, yes, Anthropic is losing money on those subscriptions, even using generously low estimates on costs².
Except to me, the argument is like a customer bringing their own plate to eat off of, and the restaurant then sues the customer and demands that no restaurant can ever be allowed to use that customers plate.
Opencode to a lot of people is a nicer and more feature rich harness than CC, it doesn’t consume any more tokens than CC, and if it did, the bounds of how many tokens each account is allowed to use is tied to the users payment and rate limits.
Anthropic has determined that they do not want to be a provider. This has clearly paid off for them, given that owning the harness allows them to do things like sign commits, which is an extraordinarily well placed advertising scheme.
Whether Claude Code can keep pace with the variety of demand remains to be seen. My guess is no and that unlike OpenAI, they are interested in suing and not buying competing harnesses.
They're perfectly happy being an API provider, where they're not selling their tokens at a loss. My guess is that they're counting the losses for Claude Code Max plans as R&D: what does usage look like if people don't have to worry about the cost of tokens? Because someday they won't, and Anthropic want to skate to where the puck will be, not where it is.
Do you have an anthropic API key with money attached to that account? Do you think most people who have a subscription to Claude code have an API key for Anthropic and have money attached to that that they’re ready to spend?
Do you think those people who have an API key are ready to just give it to some random application rather than being directed to a Web or application based sign on mechanism serve served by a company they trust?
Even if the consumers are programmers, an API key is not a consumer instrument. There is no universe where you convert a whole number percentage of users in the scenario.
> Do you have an anthropic API key with money attached to that account?
Yes, I do. Getting a Gemini API key was annoying, but getting API keys for Claude and GPT were simple and straightforward.
The web-based redirect from a command-line tool is always far weirder and more awkward than just pasting in an API key.
Since we're talking about programmers here, they should realize that both the API key and the web-based redirect end up doing the same thing: generating a token that allows this program to act as you over the API.
I don't think Anthropic has a problem with you using a regular pay-per-token API key with opencode. The issue is letting someone use their "Log in with Claude" as if it were a regular API key.
Under what law can Anthropic force OpenCode to do this? Surely it's not illegal to publish code that interacts with an API that's open for everyone to see?
The API has a very clear ToS prohibiting third-party client usage with the heavily subsidized Claude.ai subscription plans. Anthropic's right to reject or block that traffic, as well as to ban users who attempt this, is well-protected by the ToS those users neglected to read.
Regarding the legal demands here, anyone can issue anyone else a cease and desist order at any time, for anything, in the USA. The demands do not need to have merit.
"Illegal" generally refers to criminal law, not civil suits, this was essentially Anthropic threatening to file a lawsuit. Opencode was under no legal obligation to comply and was not breaking any laws, they simply decided it was easier and cheaper to comply than to fight.
Surely there's no way that's true. The logical conclusion of that would be that every random ToS is a law that everyone must abide by, regardless of whether or not they've agreed to it.
By definition, it is exactly a law. It's known as business law. The ToS is a business contract which you must agree to if you wish to use the service. Violating terms of service is literally a breach of contract.
How can you breach a contract if you are not a party to a contract? OpenCode is not using any Anthropic services, they are just publishing some source code that seems (obligatory IANAL) to be protected speech under the First Amendment [0], if this legal argument is happening in American jurisdiction.
Interesting point. I've been looking into a similar issue recently, and for example LinkedIn won a lawsuit against the analytics company hiQ because they violated their ToS for scraping their website. And I think they also never technically had a direct contract they'd breach.
Yeah good point. I think if the scraping code is written specifically for a site / system that prohibits scraping through it's ToS, the company has an edge for a lawsuit. It's a bit of a gray area I think. It depends how much of a threat you form to the company you're scraping, and how big the company is.
I think you may be confused here. Anthropic isn't going after users here at all, they essentially told another company that is interfacing with Anthropic's service in a way that violates Anthropic's TOS to "please stop or else we might have to take legal action in the future".
More broadly, you do not need to establish any kind of contractual right to "go after" anyone legally, that's not how civil law works. A cease and desist letter isn't even really legal action, it's a threat of legal action, but even then, Anthropic doesn't need your permission to sue you, just like you don't need Anthropic's permission to sue them.
If you think that inside the U.S., you have some kind of legal immunity to or protection from cease and desist letters or lawsuits from any company, for any reason¹, you would largely be mistaken. If this is important to you, you might want to talk to a lawyer.
¹ Some states have anti-SLAPP statutes that offer limited protections in certain context, but this isn't applicable in the context of this example between Anthropic and AnomalyCo.
That being said it's maybe a valid claim under "tortuous interference" theory, i.e. OpenCode damaged Anthropic by interfering with the contractual ToS agreement between Anthropic and its users
Cleary they need to get AI which has not accepted those to rewrite it. That is the easy and fast solution these days. Or at most find a person who has not accepted tos do that.
They don't need any actual written law behind their actions, all they need is money. What are you gonna do, fight them in court? Good luck with that, especially against a company directly associated with the US government and Palantir.
Can anyone explain what’s going on here? Using API is illegal? that can’t possibly be since we now know API is not even copyrighted (which personally I disagree with bit whatever)… so what is going on here?
Anthropic has two different products that are relevant here: the Claude API and Claude Code. The Claude API has usage based pricing. The more you use, the more you pay. With Claude Code, you can get a monthly subscription which gives you a fixed amount of usage. Comparing equivalent token generation between the Claude API and Claude Code, Claude Code with a subscription is much cheaper.
When it comes to third party products such as OpenClaw and OpenCode, Anthropic has made it clear those products should be using the Claude API and not the internal Claude Code APIs. OpenClaw and OpenCode have both been using the internal Claude Code APIs as when a user has a Claude Code subscription, the internal Claude Code API gives you tokens at a much cheaper rate than the Claude API. Presumably Anthropic makes Claude Code cheaper than the Claude API because they are willing to give users a discount for them to use Claude Code vs a competing product such as OpenCode.
It looks like until recently OpenCode tried to get around Anthropic's requirements by offering "plugins" in OpenCode that would allow users to use their Claude Code subscription in OpenCode. This PR mentions as much at[0][1]:
> There are plugins that allow you to use your Claude Pro/Max models with OpenCode. Anthropic explicitly prohibits this.
> Previous versions of OpenCode came bundled with these plugins but that is no longer the case as of 1.3.0
This PR seems to be in response to Anthropic threatening OpenCode with legal action if they keep using the internal Claude Code APIs.