Where Anthropic f'ed up was treating their monetization the way they treat model training. Turns out that success in experimentation is not transferrable.
They have tried to find the highest that the market pays for sota models; however, on the consumer side, this is just too confusing and unsettling:
"You can only use Fable for a week as a part of your plan" "Be ready! You have to start paying per token!" "Nevermind! we extended it for a couple more weeks" "Wait, now it's up to half your usage" "Ok, now its..."
Most people want to not care. We want our AI like electricity -- Kind of just there no matter how easy/hard is for the supply. You don't want your electricity company to be on the brink of cutting you off any second.
That's Anthropic. You don't feel they want to give you a dependable service for an, albeit premium, price. It's a constant bargaining game. That forces people to look beyond the walled garden. There, they find models that are fine... and without the shenanigans.
Yeah, I agree with this. The constant state of "...will the rug be pulled?!?" does discourage relying on it as a model and building a workflow on it. Anthropic used to just be a reliable thing you could play with. Now it's this constant source of anxiety.
It also didn't help that the government yanked it which adds another source of anxiety since OpenAI is on much better terms with the administration and the administration seems corrupt enough that they would mess with Anthropic if they got a big enough donation from OpenAI.
But anyway after Sol entered the picture, I don't think Anthropic can get away with this as much and I also think they're going to face a massive backlash from Max subscribers if they do end up ending the +50% promotion at the end of the month because Sol is a Fable peer and priced very competitively.
My wife's startup made the mistake of building her internal operations around Claude Team.
Then she hired a VA in the Philippines. Anthropic promptly banned her account without warning once the VA connected to the account. It took her weeks to get her account reinstated, at which point she had already moved on to OpenAI.
How many big tech companies let you talk to a human to get support. Automation is wonderful to cut cost for them but for the users being unable to get support is a horrible experience. But you cannot go elsewhere because they are the only player in town.
How can small companies with 1000x less money able to provide live support, but if you pay 20, 100, 200 dollars for a subscription you dont have a phone number to call ?
But Philippines is on Anthropics list of allowed countries?
As soon as I started using Fable I was like, okay, this is probably as good a model as I will need for software engineering going forward. I still feel that way. I donât need a better model, I need a faster Fable.
The thing I miss most about programming is flow, and the constant bouncing between terminal tabs sucks. Iâd love to do one thing at a time, with Fable, quickly.
remember 4 year ago we use to : have stack overflow open, documentation, obscure forums plus other tabs.
An ide open with 20 tabs open each file a component, a class or an interface We also use to hold entire codebases in our brain.
You may or may not like agents mode. I also hate flipping tabs, but I enjoy using agent mode with well named sessions. I still stick with a single session until I must move to another, then I leave them around for a few days until Iâm sure I wonât need to pick up where I left off again.
Command: claude agents
There is GPT 5.6 Sol on Cerebras if you want to try that experience for an ungodly sum of money (not getting into GPT 5.6 Sol vs Fable, but only one is available on Cerebras) for an 11x speedup.
> They have tried to find the highest that the market pays for sota models; however, on the consumer side, this is just too confusing and unsettling:
The consumer side cheap monthly plans exist for the same reason companies like Cloudflare and Vercel have a free tier: When itâs cheap and easy to get developers familiar with the tools, they will push their companies to pay the real money for those tools.
Itâs a hard balance with LLM serving because you canât really make it free. $20/month is close to free, but the $200/month plans are in a difficult place where theyâre big enough that many small companies pay for $200/month plans for their employees and ignore the enterprise features you get with the full expensive arrangements. So the companies are continually adjusting the $20-$200 plans to keep them from being reliable options for businesses, which is where the real money is.
Thereâs a short sighted cheering on of the 3rd tier and lower companies offering lower rates, but weâre already seeing them ratchet up the pricing and keep larger models closed after they get market attention.
I agree theyâve done a bit too many pricing / usage promotions and A/B tests.
The period was also marked with many billing bugs, like spending peopleâs usage credits for included Fable for a few hours (gave me a huge shock), but to their credit they refunded it.
They are fumbling the bag hard. AI's utility is for general purpose. The floor is rapidly improving from below them. With chatgpt I'm uploading all my day to day stuff. Meanwhile Claude is only for occasional super hard tech problems which are rapidly improving with being solvable easily by Sol. So what's the value add?
Their disrespect for their users is also another problem. You only get one shot to make a good impression.
> With chatgpt I'm uploading all my day to day stuff.
Same, it's been a while since I logged into Claude web.
ChatGPT web usage being separate from Codex usage limit is a nice touch unlike Claude.
Also you donât want to connect to the pipe and then after the fact find theyâve started diluting arsenic into it.
They've put themselves in a corner. Fable was too good and they gave it away with the $20 plan. It had to be a big step from Opus 4.8 to show progress, and Opus 4.8 is GREAT at coding in many different domains.
But they're getting killed on token cost. They have to get people paying more for tokens. So then they put Fable in the $200 plan and release Opus 5. I'm suspicious of Opus 5. It is mostly worse than 4.8. It _seems_ like they nerfed it to create more distance between it and Fable.
So what have most of us done? Stayed on Opus 4.8. The statistics bear this out. 4.8 still dominates.
Now they're stuck. If they take 4.8 away, everyone will riot. If they make Opus 5.x better than 4.8, they disincentivize everyone from moving to Fable and most importantly, paying more.
Really, all they can do is take the L for now and just let 4.8 be the apex of the $20 pro plan for the foreseeable future while they work like hell to make Fable THAT much better that it earns the $200 to $infinity that they really want everyone to pay.
And they've hobble Fable and Opus so hard with their safety guardrails, I ask innocuous questions and tasks and they get flagged so often I gave up on it. I can get all the work I need done in GPT5.5 or 5.6 without the hassle.
I just kept a $20 plan going for use on my phone.
They also made Fable no longer ZDR for businesses which killed tons of demand for it.
I had Fable bail on me because I used the word autopilot on a project. Changed to something unrelated, and off we went.
I havenât hit the Fable guardrails a single time after substantial usage. And Iâm working on a ML project (a game AI).
I donât doubt people are hitting it⌠shrugs
I found a crash in zsh and Fable refused to work after that.
I kind of still chill out on Opus 4.6 too. 4.8 is good too. I go between them. Opus 4.8 is a little smarter some of the time. Their use of language is both very different from Opus 5. Some of the time I have a hard time believing Opus 5 is even related to Opus 4.6 and 4.8.
I'm going to be sad when they retire 4.6. It's not my daily driver but it's still my go-to when the other models are being stupid in one way or another (either being too verbose, or lecturing me about how what I'm asking for is evil and bad).
Can you give an example? It's a bit entertaining to read this given all the years long (and still ongoing) posturing about LLM sycophancy (not that all of these couldn't be true at the same time).
Hah! I was so surprised to get that "lecturing" behavior from Opus 5 too, I didn't know it was more common.
I tested all recent Opus versions and 4.6 and 4.7 were both fine FWIW. Seems 4.8 is when something started to go wrong.
The other thing I forgot to mention about Opus 5 is that at least out of the gate, it seemed very intent on spinning up agents and obliterating my token budget. It was noticeably more token hungry than 4.8. It would make sense for them to intend this behavior.
This is likely because of your thinking level. The difference between max and ultracode is primarily that the latter is max with a bunch of agents.
Did they change the default perhaps? I didn't touch anything with thinking level in that time.
> The statistics bear this out. 4.8 still dominates.
Where can I find these stats?
The article has a chart with this data. They credit âRamp AI indexâ. The chartâs a bit confusing though, like what are the units of the y-axis?
Also on a dark reader? It says at the bottom, but gets dimmed out pretty seriously with dark reading. It's a 7 day average business spend, relative to June 1st (2025 presumably) indexed at 100.
Mythos? That's only for super special corporations, you need not apply. Fable? Can't even look at it wrong without running into cybersecurity lockouts. And even if you get past this, it's limited to <50% usage. It took me a month to complete a Fable review on my project.
So stingy. Even with OpenAI's recent usage troubles, they're still so much better than Anthropic it's not even funny. Re-ran the code review with Sol as a benchmark and it turns out Sol's performance is within 70%-90% of Fable's. Anthropic's still got the best model, but what does it matter if I can barely use it?
50% usage plus there seems to be a pretty big metering multiplier still. Do a relatively in-depth review of 3k LoC with Fable xhigh and poof, there goes 5% of the weekly Fable allowance. If I use their first party code-review skill that spawns a bunch of subagentsâwell just forget about it.
I had Max 5x and every 5h window would bite off 10% of my weekly usage. Five Fable sessions per week.
Transformer models are quickly becoming a commodity, and I suspect in time we'll all be running them locally. Even now, you can run something pretty useful on a 16 GB graphics card, and I suspect a decade into the future, entry level hardware will be running better models than high-end graphics cards can run now, as entry-level hardware gets better and models get more efficient.
It doesn't mean hosted frontier models wont exist, they'll just be rare. It's no different than any other commodity market, for example most cars are cheap commodity models, with rare individuals buying expensive luxury cars and businesses buying expensive trucks and specialized equipment.
>and I suspect a decade into the future, entry level hardware will be running better models than high-end graphics cards can run now, as entry-level hardware gets better and models get more efficient.
I doubt it. The play seems to be: lock what was once commodity compute up into datacenters depriving us regular folk of it, then sell it back to us on subscription. Even if my #NeverSubscribe movement succeeds, all that misdirected hardware [into datacenters] won't likely be practical for home use.
> The play seems to be: lock what was once commodity compute up into datacenters depriving us regular folk of it, then sell it back to us on subscription.
I had not considered that as a possibility. It is somewhat dark and unlikely, imho, but a real possibility.
I am more inclined to believe that the fab capacity will grow over time and "commodity" compute will be available to us all again.
However, I can also imagine going back to the 60s era "hyper verticalized" mainframes, in which case, the frontier labs might not just be producing models, but chips and an ecosystem around themselves and their suppliers/customers.
Yep local models will be good enough for most things you need to do, in the same way as most people need a laptop not a supercomputer.
Well nobody thought you could write software that helps tightly coordinate processes happening simultaneously from millions upon millions of nodes on nearly every corner of the world, but here we are. When industry expands further off earth, we will need more complex and intricate software to coordinate its movements, why wouldnât our systems become more powerful. If you are hopeful for humanity than you must expect the scale of industrial necessity to only ever increase alongside the imagination and capacity of its peopleâs.
For many coders including myself, LLM based coding agents work well enough to be useful, and in some cases worth paying for.
What I don't see is vast areas of industry finding $10s to $100s of billions of value in LLMs. There's no lint or compiler that can check for correctly constructed contracts. So LLMs, which should be useful to law firms, incur a lot more manual checking of their work than coding agents.
Less formal document production in other industries is likely to have less structure. That might not matter in some settings but I'm having trouble thinking of an example off the top of my head.
While there's no linter for writing contracts, my experience (as a commercial lawyer) is that frontier LLMs are far better and error checking and far quicker at writing than the average senior lawyer. The main thing holding back further deployment (in my jurisdiction) are concerns around data residency, privilege and how fundamentally it will break an industry that is so heavily reliant on time based billing.
Trouble is they're unreliable.
I compared insurance quotes last week. Needed cover for 2 brands, 1 company. Opus jumps up and down saying both brands need listing on the policy schedule. Human broker said not.
I told opus and it's the usual "thanks you're right" bollocks because it bothered to read in more detail and found that all business activities are covered.
> There's no lint or compiler that can check for correctly constructed contracts.
There are definitely linters and this exists https://catala-lang.org/
> What I don't see is vast areas of industry finding $10s to $100s of billions of value in LLMs.
Translation between languages.
That value dwarfs all programming value that can be had. Economically, culturally, scientifically, spiritually.
Between human languages? The problem will be Google is giving that away for free.
The quality of Google Translate is so low that it isn't even in the picture.
Social issues aside, that's a tiny slice of the economics of AI.
Just by nature of how often it's used etc.
You need workloads for AI be cost effective: software, automation etc.
International trade is a giant sector already, and potentially much bigger than that, now that LLM assisted translation makes it easy to offer your products and services in any language, or make complicated and sensitive deals without a common language.
Hackers always rage and down vote every time I mention this, because they are unable to see beyond their small world. Why didn't they learn that their part of the internet is 0,000000001% of what the world uses the internet for today. It's going to be the same with LLMs. Programming and hacker stuff is going to be 0,0000000000000000000000000000001% of what the world uses AI for. But translation is going to be in the top 5 of use cases.
But translations don't require anything close to SOTA level models. Translations will be high volume, low margin transactions. That will not save Anthropic.
Maybe not Anthropic, but LLM translations has a value counted in the trillions of dollars easily.
My company still hasnât been able to deploy wide access to Fable because itâs not available on a ZDR basis. This wasnât mentioned in the article but I imagine this factor is not irrelevant.
>ZDR
Zero Data Retention, for the uninitiated readers in this thread.
The no-ZDR is clearly to permit surveillance. I would be shocked if NSA wasn't all up in these SOTA model providers' systems.
Never subscribe!
ZDR for fable is coming, very soon.
Same for me, theyâre not offering it with the same data residency features of Opus, many enterprise companies canât accept compromises.
It was in one paragraph mentioned, without much detail. But yes, I agree it is one of the largest barriers.
Of course I echo the universal: Opus 5 sucks. But what also sucks is still using 4.8. Are you all seeing this? It's like the older models got dumber just before Fable and Opus 5 were coming out? I've heard the theory that it's because 4.8 is getting put on older hardware? And I imagine that just ratchets down reasoning time then possibly?
Sometimes I'm just trying to sus out if I'm truly seeing things these days or going a little nuts :)
This is yet another reason why I think local models will win in the future. They're almost certainly A/B testing all sorts of opaque stuff that people have no clue about, hence the various 'How's Claude doing this session?' popups.
So what you are paying for may vary on a day by day basis, which is quite undesirable, even if their main goal is simply to make a better model. When it comes to a tool, I'd rather have consistent mediocrity than instability.
I definitely agree. I honestly cannot do tasks that require even minor complexity. Opus 5 keeps forgetting things in context as well and coding conventions. Really cannot build with CC without Fable.
I'm not an expert by any means whatsoever but deploying models is not a straightforward task. There's a lot of levers to pull and I bet when models get "downgraded" to older hardware they do so WITHOUT the same stringent quality control of the output as they do when they release it.
I don't think it's something deliberately malicious like planned obsolescence but it's more like startup culture of "just make it fit in this sprint".
For me, Opus 5 mostly sucked because of its incomprehensible writing style. Having the system prompt focus on writing in terms that are easier to understand helped a bit
Get a daily email with the the top stories from Hacker News. No spam, unsubscribe at any time.