Anthropic officially bans using subscription auth for third party use

I'm wondering: why now, in early 2026? Why not last year? Why not in July? What changed? What does this teach us about Anthropic and what can we infer about their competition?

Does this mean that in an absurd way you can get banned if you use CodexBar https://github.com/steipete/CodexBar to keep track of your usage? It does use your credentials to fetch the usage, could they be so extreme that this would be an issue?

That's it. That's all the moat they have.

OpenAI has endorsed OAuth from 3rd party harnesses, and their limits are way higher. Use better tools (OpenCode, pi) with an arguably better model (xhigh reasoning) for longer …

I think that their main problem is that they don't have enough resources to serve too many users, so they resort to this kind of limitations to keep Claude usage under control. Otherwise I wouldn't be able to explain a commercial move that limits their offer so strongly in comparison to competitors.

I wrote a mcp bridge so that I don't have to copy and paste prompt back and forth between CLI and claude, chatgpt, grok, gemini

https://github.com/agentify-sh/desktop

Does this mean I have to remove claude now and go back to copy & pasting prompts for a subscription I am paying for ?!

wth happened to fair use ?

the subscription already has usage caps. if the caps are the caps, why does the client matter. if the caps aren't actually the caps, that's a different conversation.

Hot take: trying to restrict what front end people use to access your service is almost always an anti-competitive, anti-consumer freedom move which should be legally prohibited for those reasons. (Not just for AI, I'm talking about any and all cloud services.)

Regarding consumer freedom, I believe software running on user machines should serve the interests of the user, not the company who wrote the software or anyone else for that matter. Trying to force users to run a particular client written by your company violates this principle.

Regarding competition, forcing users to run a particular client is a form of anti competitive bundling, a naked attempt to prevent alternative clients from being able to enter the market unless they are able to build a competing backed as well. Such artificial "moats" are great for companies but harmful to consumers.

Seems fair enough really, not that I like it either, but they could easily not offer the plans and only have API pricing. Makes it make more sense to have the plans be 'the Claude Code pricing' really.

Anthropic has no authority to do as such. Users and third apps are protected by interoperability exceptions found in copyright case law.

Trying to prevent competitors from interoperating with the service also may be construed as anticompetitive behaviour.

The implementation details of an authentication process do not beget legal privileges to be a monopolist. What an absurd thought.

Their model actually doesn't have that much of a moat if at all. Their agent harness also doesn't, at least not for long. Writing an agent harness isn't that difficult. They are desperately trying to stay in power. I don´t like being a customer of this company and am investing lots of my time in moving away from them completely.

Not surprised, its the official stance by Anthropic.

I'm more surprised by people using subscription auth for OpenClaw when its officially not allowed.

This month was the first month i spent >$100 on it and it didn't feel like it was money well spent. I feel borderline scammed.

I'm just going to accept that my €15 (which with vat becomes €21) is just enough usage to automate some boring tasks.

Am I the only one perplexed why folks find this stunning or meaningful. While LLMs are novel and different in that the subscription gives you access to compute, it does not feel foreign from the subscription, free or paid, landscape. I cannot recall many (if any?) companies that would freely let you use compute, private internal APIs or anything similar just because you have a login. Maybe I come from a different era of tech but it seems both reasonable and not surprising.

Why now? It would not surprise me that this was simply an after thought and once it hit critical mass (opencode) they locked it down.

The people who they’re going to piss off the most with this are the exact people who are the least susceptible to their walled garden play. If you’re using OpenCode, you’re not going to stop using it because Anthropic tells you to; you’re just going to think ‘fuck Anthropic’, press whatever you’ve bound “switch model” to, and just continue using OpenCode. I think most power users have realized by now that Claude Code is sub-par software and probably actively holding back the models because Anthropic thinks they can’t work right without 20,000 tokens worth of system prompt (my own system prompt has around 1,000 and outperforms CC at every test I throw it at).

They’re losing the exact crowd that they want in their corner because it’s the crowd that’s far more likely to be making the decisions when companies start pivoting their workflows en-masse. Keep pissing on them and they’ll remember the wet when the time comes to decide whom to give a share from the potentially massive company’s potentially massive coffers.

In the old days, think Gmail, or before the "unlimited" marketing scam. People genuinely are smart enough to know they are doing something that they are not suppose to be doing. Even Pirating software, say Windows or Adobe. I mean who can afford those when they were young?

Things get banned, but that is OK along as they give us weeks or days to prep for alternative solution. Users ( Not Customers ) are happy with it. Too bad, the good days are over.

Somewhere along the line, no just in software but even in politics, the whole world on entitlement. They somehow believe they deserve this, what they were doing were wrong but if it is allowed in the first place they should remain allowed to do so.

Judging from account opening time and comments we can also tell the age group and which camp they are on.

At this point, are there decent alternatives to Anthropic models for coding that allow third-party usage?

It's a bit unclear to me. I'm building a system around the Claude Agent SDK. Am I allowed to use it or not? Apparently not.

Honestly seeing throttling of AI usuage across all providers:

- Google reduced AI Studio's free rate limits by 1/10th

- Perplexity imposing rate limits, card filing to continue free subscriptions

- Now Anthropic as well

There has been a false narrative that AI will get cheaper and more ubiquitous, but model providers have been stuck in a race for ever more capabilities and performance at higher costs.

I'm a bit lost on this one.

I can get a ridiculous amount of tokens in and out of something like gpt-5.2 via the API for $100.

Is this primarily about gas town and friends?

What about using claude -p as an api interface?

Too bad will stick with codex as thinker and glm5 as hands, at a fraction of the cost.

This confirms they're selling those subscriptions at a loss which is simply not sustainable.

My alt Google accounts were all banned from Gemini access. Luckily Google left my main account alone. They are all cracking down.

The reason I find this so egregious is because I don’t want to use Claude Code! It’s complete rubbish, completely sidelines security, and nobody seems to care. So I’m forced to use their slop if I want to use Claude models without getting a wallet emptying API bill? Forget it, I will use Codex or Gemini.

Claude Code is not the apex. We’re still collectively figuring out the best way to use models in software, this TOS change kills innovation.

Isn’t this flawed anyway? If an application communicates with Claude Code over ACP (like Zed), it works fine?

Instead of using SDKs, this will just shift the third party clients to use ACP to get around it - Claude Code is still under the hood but you’re using a different interface.

This all seems pretty idiotic on their part - I know why they’re trying it but it won’t work. There will always be someone working around it.

So even simple apps that are just code usage monitors are banned?

How does this impact open router?

Can’t this restriction for the time being be bypassed via -p command line flag?

You need a company with a market cap in the trillions to succeed here

Cancelled my Claude and bought GLM coding plan + Codex.

You can use Claude CLI as a relay - yes, it needs to be there -but its not that different than use the API

You guys are acting like coke addicts... dont you see?

important they have clarified that it's OK to use it for personal experimentation if you don't build a business out of it!

at least there seems to be some clarification regarding Agent SDK ... unclear whats happening with OpenClaw https://x.com/atla_/status/2024399329310511426

Why do I get the nagging suspicion their 1 million LOC codebase is backdoored?

Anthropic is just doing this out of spite. They had a real scenario to win mindshare and marketshare and they fucked up instead. They could have done what Open AI did - hired the OpenClaw/d founder. Instead, they sent him a legal notice for trademark violation. And now they're just pissed he works for their biggest competitor. Throw all tantrums you want, you're on the wrong side of this one, Anthropic.

Sonnet literally just recommended using a subscription token for openclaw. Even anthropic's own AI doesn't understand its own TOS.

So here goes my OpenClaw integration with Anthropic via OAuth… While I see their business risk I also see the onboarding path for new paying customers. I just upgraded to Max and would even consider the API if cost were controllable. I hope that Anthropic finds a smart way to communicate with customers in a constructive way and offers advice for the not so skilled OpenClaw homelabbers instead of terminating their accounts… Is anybody here from Anthropic that could pick up that message before a PR nightmare happens?

Thanks codex allows using their subscription and it’s working very well for me. I will not miss anything from Anthropic. BTW bad move, shame on you

Oh crap. I just logged into HN to ask if anyone knew of a working alternative to the Claude Code client. It's lost Claude's work multiple times in the last few days, and I'm ready to switch to a different provider. (4.6 is mildly better than 4.5, but the TUI is a deal breaker.)

So, I guess it's time to look into OpenAI Codex. Any other viable options? I have a 128GB iGPU, so maybe a local model would work for some tasks?

They really ficked up by not embracing openclaw now I use codex 5.3

People on here are acting like school children over this. It’s their product that they spent billions to make. Yet here we are complaining about why they should let you use third party products specifically made to compete against Anthropic.

You can still simply pay for API.

Codex has now caught up to Claude Opus and this is a defensive move by Anthropic

May we still use the agent sdk for our own private use with the max account? I’m a bit confused.

Just a friendly reminder also to anyone outside the US that these subscriptions cannot be used for commercial work. Check the consumer ToS when you sign up. It’s quite clear.

That's too bad, in a way it was a bit of an unofficial app store for Anthropic - I am sure they've probably looked at that and hopefully this means there's something on it's way.

Not really sure if its even feasible to enforce it unless the idea is to discourage the big players from doing it.

The entitlement from many HN posters is astounding. "Companies must provide services in the way I want billed how I want and with absolutely zero restrictions at all!" Get over yourselves. You're not that important. Don't like it. Don't use it. Seems pretty straightforward.