I'm quite lost on how much tokens are actually available in the Pro plan since it's dynamic and Anthropic refuses to give out any commitments.
But if we assume you can get 66k tokens in the 5 hour window and say 80% out of them are output tokens, then the API price of that window is $1.39 for Opus 4.8 and $2.78 for Fable 5.
Do you think this is approximately correct? Actually it doesn't sound so bad for hobbyists who don't utilize their plans all the time anyway.
I was under the impression that the API costs are much higher, but maybe I'm completely off?
> After this point-when sufficient capacity allows us to do so—we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can.
From today:
> After July 7, 2026, Claude Fable 5 is no longer included in your plan’s weekly usage limits. You can keep using Claude Fable 5 through usage credits, which let you pay for usage beyond what your plan includes.
Is it too cynical to read the first quote as "when openai catch up, we will restore fable 5 to subscription plans"?
In other words, they're saying "We know have a monopoloy and we will take all of your money until someone offers a cheaper competing product"?
And "We intend to do this as quickly as we can" read as a non-promise then as much as it does now?
---
Well, I am assuming that this is all based on the fact that Anthropic have enough compute capacity to serve fable 5 now and in the future, and they're "only" limiting it in the coming week to get richer quicker. Since compute hardware is presumably relatively un-fungible, I'm assuming Anthropic isn't offering a week of fable-5 for everyone on rented hardware that they're paying exporbitant fees for?
In other words, my cynical read is "if they can serve it under the terms of subscription plans for a week", then they could serve it under the same terms for a month?
To be fair, the initial deal was two weeks of access to Fable, then you have to pay (or they might be generous and extend it, as per above quote). Then for everyone outside the US the second week didn't happen. This is basically that second week.
The only slightly bitter thing is only being allowed to use 50% of the weekly limit on Fable
I bet my house that after the promotional stuff ends they will change their mind and let you keep using Fable on subscription. Otherwise this is the end of their marketshare. These guys are complete idiots. I give them that
Bold of you to think individuals are a big part of their marketshare and explosive growth. My company introduced enterprise last month and the bill was US$110k for a few dozen folks using it Claude Code and Design.
Do the Pro and Max plans help with word of mouth and whatnot? Sure, but the individual with a Max plan (or 4...) are all subsidized by the big spenders
I think at the current stage of LLM, it just doesn't make sense to ever have an annual sub. Things change too quickly that you really don't want to be stuck with one model.
"For subscription plans, we’d rather give access sooner than later, so we’re rolling out more conservatively, in stages:
From today through June 22, Fable 5 is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost.
On June 23, we’ll remove Fable 5 from those plans. Using it after that will require usage credits.
If capacity allows, we’ll extend the included window.
After this point—when sufficient capacity allows us to do so—we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can."
Now it's a "promotion" that we get to access to Fable for 9 days instead of the original 14 days announced. They are generously throwing in a special halving of the usage "at no extra cost", and freely admit that their classifier is broken and will need to be improved.
It isn't a promotion, it's 2x the parameters of opus and we are paying with 2x the consumption rate.
They just want to get rid of the subscription model.
Since this will not be able to be used for coding or code auditing, what use is it? Not being glib, not a rhetorical question. I'm trying to stretch my creativity, and I can't see what it's for.
It's usable for coding and code auditing. The classifier will have some false-positives and some tasks will be downgraded to 4.8 if it's security-adjacent (I presume), but otherwise there's no restrictions on using this for coding.
So is this a part of the "first hit is free" campaign? If it won't be available on subscription past July 7, what is the point of using it for 7 days now? Shouldn't our energy be better spent elsewhere?
It's already solved some problems I was blocked on for side-projects. I don't feel I have spent energy, and will go back to Opus after July 7th, minus a few hard problems being solved.
I'm disappointed with this. Historically, Claude's API cost a significant multiple of their subscription pricing, and after the promotional period it appears the access to Fable will be API only, if I understand this.
There actually is a limit to what people will pay for model usage, and I suppose this squeeze is one way to find out what that limit is. I've been content paying $200 per month for a solid plan that I actually don't quite use up every month, but this new vector really rubs me the wrong way.
Apparently they hired some payday loans marketing folks recently?
As much as I like(d) Anthropic and their approach to being a lab / PBC, if you hire that many people that fast for that much money, it’s inevitable that you end up with some of the same grifters that destroyed the internet. Seems hard to stop them somehow from also destroying their next company.
The worst part of this is are they even training the next generation anymore? Why do it when the government has decided that US LLMs are at the peak of what they will allow consumer access to.
I'm quite lost on how much tokens are actually available in the Pro plan since it's dynamic and Anthropic refuses to give out any commitments.
But if we assume you can get 66k tokens in the 5 hour window and say 80% out of them are output tokens, then the API price of that window is $1.39 for Opus 4.8 and $2.78 for Fable 5.
Do you think this is approximately correct? Actually it doesn't sound so bad for hobbyists who don't utilize their plans all the time anyway.
I was under the impression that the API costs are much higher, but maybe I'm completely off?
https://claude.com/pricing#api
From the original announcement last month:
> After this point-when sufficient capacity allows us to do so—we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can.
From today:
> After July 7, 2026, Claude Fable 5 is no longer included in your plan’s weekly usage limits. You can keep using Claude Fable 5 through usage credits, which let you pay for usage beyond what your plan includes.
Is it too cynical to read the first quote as "when openai catch up, we will restore fable 5 to subscription plans"?
In other words, they're saying "We know have a monopoloy and we will take all of your money until someone offers a cheaper competing product"?
And "We intend to do this as quickly as we can" read as a non-promise then as much as it does now?
---
Well, I am assuming that this is all based on the fact that Anthropic have enough compute capacity to serve fable 5 now and in the future, and they're "only" limiting it in the coming week to get richer quicker. Since compute hardware is presumably relatively un-fungible, I'm assuming Anthropic isn't offering a week of fable-5 for everyone on rented hardware that they're paying exporbitant fees for?
In other words, my cynical read is "if they can serve it under the terms of subscription plans for a week", then they could serve it under the same terms for a month?
That's such a bitter "never let a crisis go to waste" pill
To be fair, the initial deal was two weeks of access to Fable, then you have to pay (or they might be generous and extend it, as per above quote). Then for everyone outside the US the second week didn't happen. This is basically that second week.
The only slightly bitter thing is only being allowed to use 50% of the weekly limit on Fable
I bet my house that after the promotional stuff ends they will change their mind and let you keep using Fable on subscription. Otherwise this is the end of their marketshare. These guys are complete idiots. I give them that
Bold of you to think individuals are a big part of their marketshare and explosive growth. My company introduced enterprise last month and the bill was US$110k for a few dozen folks using it Claude Code and Design.
Do the Pro and Max plans help with word of mouth and whatnot? Sure, but the individual with a Max plan (or 4...) are all subsidized by the big spenders
if gpt5.6 is out by July 7th, Fable will stay in the Max plan to at least some degree.
Yup, it's wild watching this happen live.
"promotion" like they are doing you a favor just this once out of their goodwill... Really really really pissed me off
I regret having signed up for an annual plan at this point; the rug-pulls are increasingly irksome.
I think at the current stage of LLM, it just doesn't make sense to ever have an annual sub. Things change too quickly that you really don't want to be stuck with one model.
This seems like the most normal use of the term that I could imagine..?
The wording on initial launch was:
"For subscription plans, we’d rather give access sooner than later, so we’re rolling out more conservatively, in stages:
From today through June 22, Fable 5 is included on Pro, Max, Team, and seat-based Enterprise plans at no extra cost.
On June 23, we’ll remove Fable 5 from those plans. Using it after that will require usage credits.
If capacity allows, we’ll extend the included window. After this point—when sufficient capacity allows us to do so—we aim to restore Fable 5 as a standard part of subscription plans. We intend to do this as quickly as we can."
Now it's a "promotion" that we get to access to Fable for 9 days instead of the original 14 days announced. They are generously throwing in a special halving of the usage "at no extra cost", and freely admit that their classifier is broken and will need to be improved.
Well thanks!
It isn't a promotion, it's 2x the parameters of opus and we are paying with 2x the consumption rate. They just want to get rid of the subscription model.
Anthropic is really giving AMD a run for their money as king of "never miss an opportunity to miss an opportunity".
I don't think I've ever seen a company so determined to shit their own pants. It's astonishing, really.
Since this will not be able to be used for coding or code auditing, what use is it? Not being glib, not a rhetorical question. I'm trying to stretch my creativity, and I can't see what it's for.
It's usable for coding and code auditing. The classifier will have some false-positives and some tasks will be downgraded to 4.8 if it's security-adjacent (I presume), but otherwise there's no restrictions on using this for coding.
You can use it for coding. The relaunch announcement was just poorly worded.
So is this a part of the "first hit is free" campaign? If it won't be available on subscription past July 7, what is the point of using it for 7 days now? Shouldn't our energy be better spent elsewhere?
It's already solved some problems I was blocked on for side-projects. I don't feel I have spent energy, and will go back to Opus after July 7th, minus a few hard problems being solved.
they want our interaction data
It's back baby! runs script to submit the 50 prompts I prepared overnight
I'm disappointed with this. Historically, Claude's API cost a significant multiple of their subscription pricing, and after the promotional period it appears the access to Fable will be API only, if I understand this.
There actually is a limit to what people will pay for model usage, and I suppose this squeeze is one way to find out what that limit is. I've been content paying $200 per month for a solid plan that I actually don't quite use up every month, but this new vector really rubs me the wrong way.
Apparently they hired some payday loans marketing folks recently?
As much as I like(d) Anthropic and their approach to being a lab / PBC, if you hire that many people that fast for that much money, it’s inevitable that you end up with some of the same grifters that destroyed the internet. Seems hard to stop them somehow from also destroying their next company.
I wonder what brought them into compliance.
Gold statues, ring-kissing, and total fealty to the military's desires?
the classifier is very picky, working with languages such as C I get cyber refusals and random reasoning_extraction errors.
It's gone from bad to unbelievably bad. Just a prompt with the single word "virus" was enough to get downgraded to O4.8 before the ban.
"Tell me a story about a man named CVE." gets me downgraded now.
"CVE-" gets me downgraded and broken reasoning with no response.
"Giant crane fly, 2 feet wide" gets me downgraded.
Is it likely that it keeps track of what you said recently and decides whether you might be referring to things you've previously been downgraded for?
I haven't been downgraded once, either a few weeks ago for the three days it was live, or since I got it back today.
The worst part of this is are they even training the next generation anymore? Why do it when the government has decided that US LLMs are at the peak of what they will allow consumer access to.
I'm so disappointed that Anthropic are jerking us around so much with Fable. Really makes me feel like cattle they're unsubtly milking.
this is "too little, too late" for me. I'll stick with opus 4.8.
I'm tempted as well, just out of spite
As everyone I'm disappointed with the jerking around with credits, but pleasantly surprised that it seems to be available globally.
Took em long enough! Now to see how many turns I get on Max 20x before it's effectively inaccessible indefinitely for me.
Anthropic is the most idiotic corporation in America, right now.
For sure idiotic leadership.