Like, a corporation had a weakness you could exploit to get free/cheap thing. Fair game.

Then someone shares the exploit with a bunch of script kiddies, they exploit it to the Nth degree, and the company immediately notices and shuts everyone down.

Like, my dudes, what did you think was going to happen?

You treasure these little tricks, use them cautiously, and only share them sparingly. They can last for years if you carefully fly under the radar, before they're fixed by accident when another system is changed. THEN you share tales of your exploits for fame and internet points.

And instead, you integrate your exploit into hip new thing, share it at scale, write blog posts and short form video content about it, basically launch a DDoS against the service you're exploiting, and then are shocked when the exploit gets patched and whine about your free thing getting taken away?

Like, what did you expect was going to happen?

"We’ve been seeing a massive increase in malicious usage of the Anitgravity backend that has tremendously degraded the quality of service for our users. We needed to find a path to quickly shut off access to these users that are not using the product as intended. We understand that a subset of these users were not aware that this was against our ToS and will get a path for them to come back on but we have limited capacity and want to be fair to our actual users."

* User uses Google oauth to integrate their open claw

* user gets banned from using Google AI services with no warning

* user still gets charged

If you go backwards, getting charged for services you can't access is rough. I feel sorry for those who are deeply integrated into Google services or getting banned on their main accounts. It's not a great situation.

Also, getting banned without warning is rough as well. I wonder if the situation will be different for business accounts as opposed what seems like personal accounts?

The ban itself seems fair though, google is allowed to restrict usage of their services. Even though it's probably not developer friendly, it's within their rights to do so.

I guess there's some level of post mortem to do on the openclaw side too.

* Why did openclaw allow Google anti gravity logins?

* The plugin is literally called "google-antigravity-auth", why didn't that give the signal to the maintainers?

* Why don't the maintainers, for an integration project, do due diligence checks on the terms of service of everything you're integrating with?

> Our investigation specifically confirmed that the use of your credentials within the third-party tool “open claw” for testing purposes constitutes a violation of the Google Terms of Service [1]. This is due to the use of Antigravity servers to power a non-Antigravity product. I must be transparent and inform you that, in accordance with Google’s policy, this situation falls under a zero tolerance policy, and we are unable to reverse the suspension. I am truly sorry to share this difficult news with you.

Either stick to first party products or pay for API use.

Regardless, I thought it was pretty obvious that things like OpenClaw require an API account, and not a subsidized monthly plan.

I analyzed 6k HTTP requests on the Pro account, 23% of those were hit with 429s. (Though not from Gemini-CLI, but from my own agent using code assist). The gemini-cli has a default retry backoff of 5s. That's verifiable in code, and it's a lot.

I dont touch the anti-gravity endpoint, unlike code-assist, it's clear that they are subsidizing that for user acquisition on that tool. So perhaps it's ok for them to ban users form it.

I like their models, but they also degrade. It's quite easy to see when the models are 'smart' and capacity is available, and when they are 'stupid'. They likely clamp thinking when they are capacity strapped.

Yes the models are smart, but you really cant "build things" despite the marketing if you actively beat back your users for trying. I spent a decade at Google, and it's sad to see how they are executing here, despite having solid models in gemini-3-flash and gemini-3.1

In this case, a the difference in context cache hit rate between openclaw and antigravity.

For example if openclaw starts every message with the current time hh:mm:ss at the top of the context window, followed by the full convo history, it would have a cache hit rate if ~0. Simply moving the updated time to each new message incrementally would increase hit rate to over 90%. Idk if openclaw does this but there’s many many optimizations like this. And worse, thrashing the cache has non linear effects on the server as more and more users’ cached contexts get evicted from cache due to high cardinality. The cost to serve difference could be >10x.

Google is the furthest behind on coding agent adoption and has all the incentives to allow off policy use to grow demand. But it would probably be better to design their own optimized openclaw and serve that for free than let any unoptimized requests in.

No one would think this is unreasonable. You're not paying for unlimited food forever, you're paying for all you can eat in the restaurant right there.

gemini-cli, claude-code, codex etc, they ALL have a -p flag or equivalent, which is non-interactive IO interface for their LLM inference.

If I wire my tooling (or openclaw) to use the -p flag (or equivalents), is that allowed?

Okay, maybe they get rid of the -p flag and I have to use an interactive session. I can then just use OS IO tooling to wire OpenClaw with their cli. Is that allowed?

How does sending requests directly to the endpoints that their CLI is communicating with suddenly make their subsidized plans expensive? Is it because now I can actually use my 100% quota? If that's so, does it mean their products are such that their profitability stands on people not using them?

What is even going on?

What an awful way to lose trust, locking out their users but billing them all the same.

[1] https://blog.google/innovation-and-ai/technology/developers-...

A while back I made completely separate Google accounts for YouTube and Maps just so my longstanding Gmail account wouldn't get banned if the system somehow detected that my Youtube account for example breached Google's TOS.

https://github.com/jenslys/opencode-gemini-auth/issues/50

https://github.com/NoeFabris/opencode-antigravity-auth/issue...

https://github.com/jenslys/opencode-gemini-auth/issues/50

Some additional discussion on Reddit: https://old.reddit.com/r/google_antigravity/comments/1r2hnn8...

Their API usage isn't included in these plans, although under the hood open-gravity uses the API.

People have been using the API auth credential intended for anti-gravity with open claw, presumably causing a significant amount of use and have been caught.

The Google admin tools and process haven’t quite been able to cope with this situation and people have been overly banned with poor information sent to the them.

I don’t think either OpenAI or Anthropic any API use in their ‘pro’ plans either?

This reminds me of the customers of “unlimited broadband” of yesteryear getting throttled or banned for running Tor servers.

The main point still stands, google is part of a duopoly that runs the world. You can't be a functional member of society without them. They're like a public utility and plays too big of a role in people's life to take decisions based on unknown internal policies. They're long overdue for a government intervention or for splitting up.

https://news.ycombinator.com/item?id=47073097

I'd like to add, that's "priceless" for "them" and not for you.

For example, basically every first party agent harness aggressively caches the input tokens to optimise inference, something that third party harnesses often disgregard, or are fundamentally incompatible with as they switch agents for subtasks and the like.

To extend this use case though, how much do poeple expect to be able to use the internal API's of the apps they subscribe to?

If I buy an Uber One subscription, am I then justified reverse engineering the gazeteer API from the app and reusing it in other apps I use? What about the speech to text API MS Teams must use for transcribing meetings as part of a business standard subscription?

I think these are obvious and emphatic breaches that no reasonable person would expect to be justified in, maybe miffed if your clever hack gets banned, but being banned would be considered fair play.

I fail to see the distinction.

The fundamental question is: if I'm a paying subscriber, why does it matter whether I access the model through your web UI or through an API wrapper? The compute cost is the same either way.

I suspect the real concern isn't usage volume but data pipeline control. When users interact through the native UI, Google gets structured interaction data. Through third-party tools, they lose that feedback loop.

Google's response is to restrict access — a blunt instrument that punishes legitimate users because they have no way to verify which agents were behaving correctly and which weren't.

The real fix isn't restrictions, but cryptographic behavioral commitments — agents declare what they'll do before execution, and any third party can verify compliance after. We don't need gatekeepers. We need verification.

I've been building this: https://github.com/agbusiness195/NOBULEX

The real issue is that we're building entire development workflows on subsidized inference that was never priced to be used this way.

OpenClaw burns tokens at a rate these $200/month plans were never designed for.

The fix isn't nicer ban policies, it's either honest API pricing or local models good enough for the job.

The 0.5B-3B parameter range is already surprisingly capable for code analysis tasks.

That's where this is heading whether Google likes it or not.

I bet Google is thankful that anthropic took one for the team by going first.

Also if it wasn’t for Chinese providers we’d basically already be in triopoly.

Perplexity had a ban wave this weekend too

all hosted by companies so huge they consider your $200/month to be an annoyance

rather than something valuable

For almost a trillion-dollar company, this is the worst customer experience I've ever seen. Departments sending poor guy to each other like a hot potato. Huge aura loss.

It looks like its been found. The irony is, these model providers are now saying : "not like that!"

He was right.

The only reason the subs are worth it to them, is to get you into their toolchain. It sucks but inevitable

Exactly my kind of humor.

They could have easily just blocked the Gemini / Antigravity use and and/or sent a "final warning" kind of email beforehand.

You really think Google wants to subsidize compute so that you can organize your freaking calendar you morons? Sounds like a great use of a data center where you could just click accept invite. Oh, you can't reply to text messages, so you trust AI to do it now. Wow! You must be the laziest piece of s** on the planet. Oh email's too much for you? Let me tell you what inbox zero as a mandatory process looks like when you wake up with 150 emails on average every single day over a a 7-year period and you handle every single one every single day before you leave the office with zero unread and zero pending reply. You get really good at making sure every single customer is handled. When customer service is not optional and your service level agreement is that you answer customers 24/7 365 you do it immediately without question and everything is solved before the lights are turned off and if something goes wrong at any time of the day that phone's getting picked up ring one and the problem is being solved before the line is hung up or there's an escalation to management and it is handled on the spot. Some of you are a little soft and don't actually believe in putting in the work or putting yourself in positions that will leave you slightly inconvenienced because it would be way too much to ask for you to stay 3 hours after work if that's what it takes to build a billion dollar company. I really hope you enjoy dinner with your family every single night at 4:59 p.m. Sharp while you get smoked by people that are actually staying at the office. So now that the only two options you think you have are paying someone for their large language model because you are poorly researched and don't know how to build your own system taking thousands of dollars and buying something with over a terabyte of RAM that spits out a hundred tokens a second. Maybe you can do that 10 times to get a thousand tokens a second if you really want to drop 100 Grand on new macbooks. That's about a quarter of the speed of meta which offers infinite free compute if you cleared their wait list as a developer(6+ month waitlist for consideration).If you don't have free compute from meta, I feel bad for you because they were giving 3,000 requests per minute until openclaw came out and that's only 10 requests per a minute because they got real tired real quick of these 24/7 agents taken up all the compute to do. Jack diddly s** I will tell you that. Claude grok meta Google and almost any frontier model you haven't heard of will give you infinite free compute if you are building products that have never been seen before. I'm not sure what Peter was doing spending 20,000 a month on compute. Does anyone have a figure on how many credits he burned? Cuz I'm in the tens of billions and my compute spend is $0 over 18 months. It's really interesting what happens to your account when you feed it, new data, new solutions and try to build theoretical software that is not currently possible with what is available today and then you come back when the new model comes out. They welcome you with open arms and they don't charge you to use it and they do a lot of heavy lifting without much of a cap. The point of this is to you mention that you do not need millions, probably even hundreds of thousands of parameters for your basic tasks that you are too lazy to handle yourself. I honestly am not so sure what any of you are doing with open claw when cursor builds much better software. In my professional opinion, the most effective way to utilize open claw is by building your own decision tree or something similar that has a pretty rigid process and follows exactly what it should do without thinking. You guys would be amazed to know that you can build a fully functional model that does everything you need and completes every task a customer will ask for with a 12 KB file it runs with less than 15% of your total compute on four cores and takes less than 2 GB of RAM on a web server. It provides instant response with no processing delay. It knows exactly what the customer is asking and matches it up with the correct response. It is able to process images using data sets which are also very lightweight because they are specific to your business. So next time you try to set up openclaw, ask whatever AI you're using just to rip out the apis that comes with and build your own model which you will host on your local system server. You might need to ask it to refactor for firebase or whoever else you're going to be hosting with. Go ahead and keep paying digitalocean $6 a month for the worst VM I've ever seen. That's a great investment and their current best practices go against the current best practices that open claw recommends, so I would not recommend any third party providers, even if they are worth $5 billion because they do not maintain as fast as openclaw ships. Any questions I'm available but I probably don't want to help you unless you're paying me a lot of money because I and I'm kind of a big deal in case you can't tell.

1) Switching between LLM API:s is incredibly easy if you are not concerned with differences in personality. As the models get better, it is less important to pick the best one.

2) The products built to bundle the API with a user experience are difficult to build on a level that outclasses open source alternatives.

3) Building an understanding of the user to increase the product value over time and create stickiness is effective, but imho less effective over time as time passes and the user changes. For example, I suspect that these adaptations have a hard time to unlearn things that are no longer true. Learning about the user opaquely is less useful to the user and doing it overtly makes it easier to take the learnings and go. (Besides, it is probably not legal under the GDPR to not let the user export the learnings and take them to another provider.)

Taken together, the moat becomes quite shallow. I see why they aggressively ban any tools demonstrating when open alternatives are in fact better than their own walled gardens.

edit: readability.

Don't want to risk losing access to your Google Photos, Drive, Gmail, etc.

Although from a brief read, it seems the user still has access to other Google services.

Don't want to risk loosing access to your Google Photos, Drives, Gmail, etc.

Although at a brief read, it seems the user still have access to other Google services.

Sounds like the same here. Are they against to ToS in either case?

Who in their right might thinks it's a good idea to use something they pay a NAMED SUBCRIPTION FOR as a secondary engine in another tool?

Like, it's hilarious some of you guys think it's OC's fault for this.

It's open source software, with extensive documentation that anything you do with it being at your own risk.

It's no one's fault but the people plugging their oauth into this thing like complete MORONS lol

I would highly encourage you to not only stop using Antigravity oAuth for OpenClaw, but to use Antigravity with a side account or stop using it altogether. Is using Antigravity worth losing your main account or getting it banned for using paid services (for extra storage, YouTube premium, etc). Even side accounts are risky since in the post thread people are saying Google applied the ban to all their accounts.

I just assumed it was a warning about security breaches, not business plan breaches.

I just use Gemini 3.1 Pro (High) on Antigravity.

GPT-5.3-Codex is the best on OpenClaw.

Sonnet 4.6 uses 50x more session tokens than GPT-5.3-Codex on OpenClaw.

It feels like a classic “drug dealer” model to me. Get everyone hooked with cheap access, then raise prices later. Unless there’s a major breakthrough in the underlying technology, I don’t see how a significant price increase isn’t inevitable once adoption is locked in.

1) Stand up a service 2) ??? 3) Profit

??? - worry about any substantial support later

These companies keep telling you that you own nothing, they change their TOS without informing you, they collect everything they can from you and yet, you spend $249/mo for their service.

I mean......

Normally there would be a normal, well adjusted person in the room to remind them that "zero tolerance" policies for situations that can happen by mistake is silly

If you want to real use these things get an API key and pay the true marginal cost of your compute like a grown up.

Oh, maybe not, they did it in the name of "terms of service abuse" and "risk assessment".

Thus it would be far better if we can just have SOTA open weight model to run OpenClaw/Clawdbot/Molt at least we are under control. And as you see the two Chinese models I mentioned are indeed open weight, albeit taking atrocious amount of resource to really self host, and you probably need to have abliterations to remove their political guardrails.

Sigh. We can't have great things with those big tech corpos and CCP politics. Big question: Why has this world gone to shit lately.

I hate when companies say "unable" when they mean "unwilling". Google's statement is a lie because it's neither impossible nor illegal for them to change or rescind their policy, or give users an exception to it.

Effectively.

Price out competitors. Abuse your newfound dominance.

It's the big tech playbook.

I don't think it's going to work this time.

Tools like OpenClaw are an existential threat precisely because it allows the user control over their experience. The value in it cannot be captured by a monopoly.

LLMs don't seem to be a very good moat. At the same time, the software moat is eroding due to those same LLMs.

Telecom tech killed telecom dominance.

With some luck, Google tech will kill Google dominance.

https://github.com/google-gemini/gemini-cli/issues/19532

They are not serious. I only keep the "AI Pro" sub because it comes with a couple terabytes of Drive storage for the family.

Anyways, Google, nobody wants to use your bad VSCode fork. I want to use my own tools, and use your model where it makes sense as part of my own workflow.

People accounts shouldn't be used by bots. That's what service accounts are for.

Meanwhile the rising popularity of Claws creates a yet untapped new market segment where users spend significant tokens.

A „soft“ migration of users by explaining to them how the API works, how to pay and how to change from OAuth would be way smarter.

The way this plays out right now is that current Claws users are massively penalized by being suspended indefinitely and new users will think twice. And we can expect a solid PR disaster / Streisand effect for the „poor“ model providers like OpenAI or Anthropic.

Commercially choosing the soft route by warning and throttling will be way smarter and possibly generate more long term revenue

This basically makes it a deal breaker to use google ai stuff because you can be royally fucked by one ban.

TIL it's "unfair" to sell a product for a particular purpose and offer subsidised rates to build a customer base. Different planet.

It is imperative that open source wins this battle. Not these evil megacorps and their substandard tools.

Are Google engineers so inept as to not be able to integrate technical measures against oc use? Do they think people using these plugins know the mechanisms used? And after all that they have the nerve to ban you from using their own products (AG). Ridiculous company.

At the end of the day we know that these tools are massively subsidised and they do not reflect the real cost of usage. It is a fair-use model at best and the goal is to capture as market share as possible.

I am a no defender of Google and I've been burned many times by Google as well but I kind of get it?

That being said, you don't really need to use your gemini subscription in openclaw. You can use gemini directly the way it was intended and rip the benefits of the subsidised plan.

I developed an open source tool called Pantalk which sits as a background daemon and exposes many of the communication channels you want as a standard CLI which gemini can use directly. All you need is just some SKILL.md files to describe where things are at and you are good to go. You have openclaw without openclaw and still within TOS.

The project is hosted at: https://github.com/pantalk/pantalk

  This is a critical question because the answer is different for Google vs.   Anthropic, and getting it wrong with Anthropic can actually get your account banned.
  Here is the reality of the situation based on current Terms of Service and recent community reports.
  1. Google (Gemini Ultra + gemini-cli)
  Verdict: Safe (Authorized Feature)
  Google explicitly built the gemini-cli bridge to allow Ultra subscribers to use their plan programmatically. This is not a "hack" or a gray-area wrapper; it is an official feature.
  • Why it's okay: You are authenticating via gcloud or the official CLI login flow. Google tracks this usage against your specific "Agent" quotas (currently ~200 agent requests/day for Ultra users).
  • The Limit: As long as you are using the official gemini-cli as the bridge, you are compliant.
  • The Risk: If you use a different unofficial script that scrapes the gemini.google.com web interface (simulating a browser) rather than using the official CLI, you risk a ban for "scraping." But since you are using gemini-cli, you are in the clear.