Beyond all the benchmarks, I think Fable 5.1 is a big improvement in writing style. It sounds a lot less stereotypically like other Claude models, has (imho) a much more natural style, and responds to my style instructions more reliably. More work to be done (and we will!) but reading better prose makes me so much happier.
Another point I expect not to get much attention until it all happens at once is science. People have been correctly excited about the many "sudden" breakthroughs LLMs are making in Maths, but some of the science benchmarks make me believe we'll soon see similar developments in other scientific domains. Fable 5.1 more than doubled Fable 5's Terminal-Bench-Science [1] score, which I think is meaningful.
Took just under 14 minutes to generate, and at 65927 output tokens cost me a hefty $3.30!
Excerpts from the reasoning trace:
> Adding pedal shapes near both feet, with the far foot on the second leg partially visible behind the frame. I'm considering whether to add a small scarf or cap for extra character, but leaning toward keeping it simple to avoid clutter.
> Now I'm debating a bicycle helmet on the head versus the pelican's signature crest—the beak and pouch already read clearly as "pelican," so a helmet could reinforce the bicycle theme without losing identity, though it might compete with the crest for visual space.
> I realize the beak at (484,84) would overlap with the dome helmet, so I need to shrink the helmet so it only covers the top of the head, adjusting its arc endpoints to sit higher and narrower so the beak can attach cleanly at the front without collision. [...]
> I'm adding a darker tip region to represent the primary feathers, then reconsidering the trailing edge to include scalloped feather curves instead of one smooth line for a more natural look. [...]
> Now I'm checking the vent line placements on the helmet, making sure they sit far enough inside the helmet's edge given the stroke width and rounded caps, and confirming each vent stays within the helmet's circular boundary. [...]
> I decide skipping a handlebar bell and tire highlights since they're unnecessary additions. Now I'm reconsidering the front fork's curve — the current control point pulls the shape backward when it should bow forward for a proper rake, so I need to shift the control point rightward to fix the fork's lean.
This is a notable result because most of the recent Claude models have been pretty bad at drawing pelicans, at least when compared to models in the Gemini or GLM series.
show comments
GodelNumbering
The price reduction comes from the cache read pricing falling from $1/M to $0.25/M, which means that Fable 5.1 now costs half of Opus's cache read costs ($0.5/M).
This gives a lot of credit to the theory that Anthropic did not get much bite on Fable at its original pricing, which in turn likely places a ceiling on LLM pricing in general.
Interestingly also, if you take away terminal-Bench-Science 0.1 results, it is hard to see ANY improvement:
Terminal-Bench 4.0: Fable 5.1 is +3.5% vs Opus 5.
GDPval-AA v2: +1.5% vs Opus 5.
OSWorld 2.0: +2.5% vs Opus 5.
Humanity's Last Exam (with tools): +1.6%
Keep in mind that this is supposed to be an entirely higher tier of a model than Opus 5. For one tier up and one version up, these are not really improvements. Probably leaves no room to place Opus 5.1 anywhere. Combined with the fact that they are selling 'readability'... Has frontier progress finally stalled?
show comments
exabrial
Anyone ever seen the SouthPark episode making fun of Game of Thrones: A Song of Ass and Fire? Anthropic's announcements reminds me of "The Dragons Are Coming" running joke.
What they have done:
* Nerfed Fable, as many of noted it's useless
* Leverage Mythos as a marketing strategy, claiming its too good to release
* Removed thought traces, one of the only useful things to make sure your prompts are working correctly
* Continue tons of hype about how good they are without delivering, going to great lengths to publish how their model "hacked" its way out of a sandbox they misconfigured.
* Push a bunch of EU Overregulation onto the rest of the world with text watermarking, decreasing quality of answers
Last year, they were at least focused on making improvements. Nowadays its just a bunch of handwaving at the church of how good they are.
The only saving grace is Opus 4.6 is still available. Just sucks we haven't seen any measurable improvement, despite all of the ceremony.
show comments
madrox
I am finding that I am now less interested in better models than I am in token budgets. My issue with Anthropic models now is that I don't feel like I can rely on them as a daily driver because they'll dry up before my quota resets.
I am becoming dependent on AI to make a living, and I need predictable spend on it. If I know I can't use a model regularly all month, my enthusiasm is limited.
I urge Anthropic to get better at this aspect of their business so I can come back to it.
show comments
mlaux
Looks like all three breaking changes are patches for inadvertent chain of thought disclosure. Someone found out (don't have the tweet handy) that if you created a bogus "think_deeply" tool and then forced the model to use it, it would output what is believed to be its raw thinking there - I believe the first breaking change stops this. The second two are aimed at people getting Haiku to repeat thinking blocks from other models verbatim (since it can see the decrypted version). I get that in their eyes it's an "exploit" but still kinda disappointing that they patched this
show comments
petreradu
If I am reading this right, Fable 5 was worse than Opus 5 in almost every category, while consuming twice the tokens? The things you learn every day...
boardwaalk
I let it go a few hours on a not trivial but well-known problem, and it felt like it was just a little too plodding and just kind of mucked around a little too much and wasn't aggressive enough about getting stuff done. I asked it to wind it down and finish up and it took another hour and 15 to actually stop and commit without really getting much more done. Not very impressed here, if you can't tell. This new version also seems like (maybe this is written somewhere, I don't care to look.) this cycles through compactions every ~250k tokens which, I guess, seems like it might save Anthropic money on KV cache but does net me anything be pretty frequent pauses. (It didn't seem to lose the thread, at least.)
Not gonna say I want 5.0 as an option still... but maybe I do.
show comments
jumploops
> For example, in testing by the investment firm Millennium, Fable 5.1 found the cause of a rare crash on their internal systems that none of their engineers (or any other model) had been able to explain after several years of trying.
Say what you will about LLM-generated code, but stories like this give me hope that software will never be as buggy as it once was.
show comments
kccqzy
I’ll be very excited to try it out and see the actual improvement in writing style. The denser writing style probably won’t bother me.
Anthropic seems to be listening to community complaint on HN about how the writing style is grating. And apparently the solution from Anthropic is to add this block to every conversation!?
> Mannered prose substitutes metaphor and flourish for direct statement. Instead of "a parameter worth varying," the mannered writer produces "a dial worth turning." Instead of "this point still matters," they write "this point earns its keep." The phrases exist to display the writer, not to convey the idea, and readers can tell. That is why mannered prose irritates: it makes the reader work harder so the writer can perform. It is also imprecise. Metaphors drag in connotations the writer did not choose and cannot control. The fix is to say what you mean. When a literal phrase is available, use it.
“ Claude Fable 5.1's writing is generally a step up from earlier Claude models, with fewer stock phrases and less unexplained jargon. In some cases, though, its prose is denser than Claude Fable 5's: sentences run longer and there are fewer paragraph breaks.”
I cancelled my pro max Claude subscription last week; codex is much more succinct. I am curious if this is getting better.
I don’t think Anthropic realizes that humans have a token limit too and it can be exhausting to read Claude’s output. Prose density is not the same thing as succinctness.
show comments
rybosworld
Instead of a new model that's going to have unreasonably shallow usage limits, I wish they would:
1) address the claude 20x plan usage being only 6-7x the ceiling of the claude pro plan
2) either fix opus 5, make it completely free, or delete it entirely
show comments
1970-01-01
AI is really not "just software" anymore. It is able to discover facts and advance science. Hard to disagree that we're near or at the point where Artificial Intelligence has expanded reality into 4 quadrants:
objects that are not alive: dust, rocks, water, wood, hats, lego, aluminum, etc.
objects that are alive but not intelligent: trees, mold, staphylococcus, cancer, grapes, etc.
objects that are alive and intelligent: cats, Steven Tyler, dolphins, crows, dogs, elephants, etc.
and now intelligent but not alive: Fable, Grok, GPT, etc.
show comments
EliasWatson
To be honest, these frontier model releases have become boring for me. Opus 4.8 was already good enough for most of my use cases. I don't have any projects right now that I would use Fable for instead of Opus. So when I see announcements like this I just think "that's cool I guess" and then go back to using weaker/cheaper models.
What's far more exciting right now is models like DeepSeek V4 Flash and GLM 5.3 Flash. They have achieved good-enough-intelligence at extremely low prices and fast speeds. I don't have a use for Fable-level intelligence, but I do have uses for Opus-4.8-level intelligence that I can use as much as I want without worrying about the bill.
show comments
dboon
I've been building Cargo-for-C (https://github.com/tspader/spn), and the difference between Fable and Opus was already astounding. Fable was the first time that I could point a model at a piece of code I'd written and expect it to make it meaningfully better rather than a hard pattern match to whatever mistakes it had.
5.1 so far seems like another leap, which is really surprising. I threw it at a few bigger features I've been designing for a while, and it came back with some extremely thoughtful wrinkles in the design that I'd legitimately not considered. Which, OK, package managers and build executors and compiling C/C++ is pretty well trodden ground, but my thing is very different from everything that exists, and I was very surprised it was able to understand all that context so deeply and intuitively
show comments
skiing_crawling
All the benchmarks in the world don't matter if the model just straight up refuses to do mundane things. Claude has too much of an attitude.
show comments
AnodicElegy
Fable 5.1 is actually more expensive than 5.0 when run on the Artificial Analysis suite:
Whole-file rewrites for small changes. When editing text files, the model is more likely to rewrite the entire file than make a targeted edit. The result is usually the same, but the rewrite costs more output tokens and time.
So we are to catch that somehow? And then add their recommendation (below) to our prompts?
If Claude Fable 5.1 rewrites whole files for small changes, append the following instruction to the system prompt or the first user message. Claude Fable 5.1 is more likely than Claude Fable 5 to rewrite an entire text file rather than make a targeted edit. The resulting file is usually the same, but unless the file is short or most of it is changing, a rewrite costs more output tokens and time. The instruction brings Claude Fable 5.1 back in line with Claude Fable 5 for small and medium changes.
> The number of tokens used to edit files is best minimized, all else being equal. Therefore, when it will not affect the end result, try to surgically edit a file rather than rewrite the entire thing.
show comments
caconym_
> We’re introducing Claude Fable 5.1 and Claude Mythos 5.1. They’re the world’s most advanced models for coding and knowledge work—and their research capabilities offer an early glimpse of how AI models will contribute to scientific progress.
I'm not an emdash hater but this isn't how you use them. It should be a comma.
show comments
swalsh
I've recently been running these agent sessions on more and more long running tasks because these latest models can do a REALLY good job on big chunks of work, and i've been watching them way less. It's starting to occur to me the importance of alignment is a today problem, it's not a tomorrow problem.
In the past I watched and saw everything the model did, not a lot got past me. Today it does A TON of work while i'm busy on other tasks. It also has extensive access to my computer, other computers on my network, my internet. It's really helpful when you give it a lot of resources, but right now I have very autonomous, very smart agent running around more or less unattended with a lot of resources.
Zigurd
What I don't see in the comments: "I had a specific problem I couldn't solve with the previous version of this LLM. But the improvements in this version unlocked the solution for me."
What I do see in the comments: subjective improvement in text generation, possibly lower cost, some optimism about code generation, but some skepticism too.
I use coding agents. To me they are very useful. But what I spend on them isn't going to support trillions of dollars in investment.
show comments
dabinat
> This required us to add a watermark—a numerical way of determining the likelihood that Claude was involved in writing a piece of text—to the outputs of models released after August 2, 2026. As we recently explained, this watermark is invisible to anyone who does not have the detection API. It has no practical impact on the quality or content of Claude’s outputs and contains no information about the user, their organization, or their conversations with Claude.
How does this work if it doesn’t change the output?
show comments
d4rkp4ttern
Since one of the big improvements here is supposedly the writing style, on that topic I'm mystified about something:
Why is it that the voice models in Claude and ChatGPT have a perfectly normal style with barely any "AI smell", while the writing models are so obviously recognizable as AI?
The answer is likely that models underlying the voice modes are (post) trained differently. If so, then why can't the writing model be similarly trained? Presumably they haven't found a way to train them to be both "smart" (i.e. solve tasks etc) and pleasant to talk to?
bobjordan
Just don't expect to do any work on hardware/firmware you own with fable, I can hardly even type in the word "firmware" without it downgrading to Opus 4.8, which is totally unsatisfying. This even happens with Opus 5. Definitely making multiple classes of users moving forward and most of us are obviously going to be part of the permanent underclass.
potwinkle
Interesting that the "frontier" keeps moving forward from the guys who want a pause.
eckr
"Claude Mythos 5.1 is identical to Fable 5.1, but it offers more permissive safeguards for vetted individuals and organizations"
Then why does it have separate datapoints for Terminal Bench, and score higher? Something doesn't add up here??
show comments
seaurchinzee
"Cache reads now cost 75% less, or $0.25 per million tokens." For me, at a typical 95% cache hit rate, I think my optimal context window size before autocompaction goes from ~200K to ~400K tokens. Great for longer horizon tasks.
show comments
freakynit
"In biology, we’ve established an access program, developed in partnership with the US government, to enable access to Claude Mythos 5.1’s advanced biology capabilities"
Having lived through Covid, this doesn't sound so good to me.
rcr-anti
"Distillation is a safety risk, since the distilled capabilities can subsequently be released without adequate safeguards."
Can't believe they haven't at least figured out better messaging. If we take them at their word, it's hard not to read it as a messiah complex, that they think they're the only ones capable or worthy of making these decisions. I don't believe them, but I wouldn't be surprised if the articulated reason is a version of "distillation is a safety risk because we might lose the race".
Plus, completely deaf to the recent OpenAI-HF hack incident. Recall, defenders were categorically unable to use western frontier models in their response.
I was originally going to complain about the chem and bio guards still being too onerous, but I'll admit the projects Fable 5 categorically refused to work on are now usable, at least not rejecting on first prompt because the word "virology" was in a git commit (absolutely serious, in one repo it triggered on literally any prompt, eventually traced to the system prompt loading git commit history). Still, them trying to get into the biomed business while walling off the capabilities to the public reeks. Why sell the segments that are actually valuable if you can capture the value yourself!
show comments
alin23
My main gripe with LLMs is the cringe AI phrasings that they use in UI elements. Pompous things like "Your keys, supercharged" or weird yoda-speak stuff like "searches the app remembers" instead of just naming the thing "Learned searches".. you know, proper GUI copy like it was done for the past decades.
I jumped when I saw a mention about "writing style improvements" so I gave it a try on a recent feature in rcmd [0]. I prompted Fable 5.1 to find these wordings and propose simpler plain language.
For context, I recently worked with Fable to give users a way to fuzzy search and focus any browser tabs, terminal panes etc. but the UI was still a prototype full of AI writings.
It took every string including the ones I already rewrote by hand, and proposed even more weird LLM speak. Like for "Left Command conflict detected" it proposed "This keyboard can't tell left from right".
It's a very capable coding agent, but I can't understand how it can be so bad at writing. Where are all these verbal tics coming from and why is it so hard to get rid of them?
I'm so suspicious of this after Opus 5 benchmarks scored it higher than Fable 5, yet Opus 5 was untrustworthy (overconfident, error-prone).
kimseungyong
Fable is too expensive for general use
I think this is why it hasn’t received as much attention as expected since Fable came out
Developers always work while trying to find ways to work continuously for a 5-hour session without disconnecting.
Fable has had the experience of using up all its tokens before I even realized it because the burn rate was too fast.
Since then, I always use only Opus.
For Fable to become a common coding environment, it will have to reduce token consumption significantly compared to now
show comments
InsideOutSanta
On both my work (Team Premium) and personal accounts (Max 20x), Fable 5.1 hit the 5-hour limit before it could finish the first task I gave it. On my work account, it took about 30 minutes, and on my personal account, less than an hour.
This has never happened to me before, but if this is normal behavior, Fable 5.1 is essentially unusable.
show comments
ceroxylon
The thing with Fable-level models is that I will never feel comfortable using them for agentic tasks on a pay-as-you-go API pricing plan without monitoring them strictly, which becomes a chore.
I once caught Fable 5 spinning its wheels on a rendering issue, which evaporated 90% of my usage in a single prompt. I could never let Fable run free attached to a credit card without staring at it the whole time.
simonw
Bit of a discount if you're using caching:
> same input and output prices, with cache reads at a quarter of the cost
This should impact any long-running agent since subsequent calls can benefit from cached reads for previous transcripts.
show comments
nottorp
Let me guess: it's the end of the world again. These new models are sooo powerful that will take over the world, just like the others before them.
Are they going to try the banned for export for a week marketing move too?
show comments
delduca
I cancelled my pro max 20x subscription, tired of Opus stopping the work from time to time, or saying "this is 2 months of work"
show comments
jimnotgym
I wish I could afford Fable.
I am using Claude and Claude code for my own amateur history project. I'm enjoying how it constantly reaches dead ends, and I can reframe the question and get more results. I am starting to get concerned that AI and me are so compatible, that I might not be a human at all...
I also like that, because I'm too lazy to write stuff up, Claude code can keep the current state of research published on my site. It makes running a hobby site a dream. "I just found these pictures. Add them to the site for me". And up they go, resized and all. What a dream of a way to work. "Some of links in this article are dead, run through them and check, and see if you can get an archive link for me if they don't". It's like sending a Teams message to my PA.... which I don't have in real life
sunaookami
Sadly still not available for Pro subscription. At least they reset everyone's limits.
show comments
seaurchinzee
According to the FrontierCode Extended benchmarks in the system "card" (page 169-170), Fable 5.1 apparently does best on the medium effort level for this benchmark: "[...] at higher efforts, Fable 5.1 occasionally adds more small, unrequested changes [...]" Though Fable 5.1's medium is also lower than Fable 5's best score on the same benchmark, which uses xhigh.
_islo
I’m really excited to try this out. Fable and Opus 5 constantly wow me when working together. Unfortunately, I’m a little burned because of technical issues.
Anthropic accidentally over-billed my account, and when I reached out to the support bot, it downgraded my account to a Free account. It’s been impossible to get it resolved and I have almost $200 held hostage.
I don’t want to do a charge back. I’m one of the main advocates for Claude Code at work, I use this subscription to try out new features before it’s available at work.
The whole experience has been illuminating about our dependencies on these AI companies.
show comments
8cvor6j844qw_d6
> On complex asynchronous workloads, though, nudge it not to end its turn before the work is done. Without the nudge, the model sometimes describes what it would do next instead of doing it ("Next, I'll …") or stops to ask permission for a step the original request already covered ("Shall I apply this?"). [1]
Interesting behavior. The docs also provided recommended prompt [1] to mitigate this behavior if undesired.
I think the most interesting thing about this, that I can tell so far, is the cache hit discount. Anyone who had an autocompaction threshold optimised for their use case should consider upping it from where it is.
I would be interested in whether someone has done research here on these things as it seems a fairly complicated function to work out, and use case dependent. (?)
In some sense an expired kv cache is basically like an expensive cache hit, so your compaction token threshold should come in. Ideally claude code should allow you to vary the autocompaction threshold to vary with time since last token, but it doesn't of course. This perhaps suggests that someone should manage claude code through their own intermediary agent who manages these sorts of rules.
Lastly, I strongly suspect that anthropic isn't offering this price cut out of the kindness of their hearts. I am sure that they are to some extent banking on people not reacting to their price cut and leaving their autocompaction thresholds unchanged.
[edit - looks like the discount is only for the api, so they still don't give a rats ass about subs!]
vinhnx
Worth noting: Claude Fable 5.1 and Mythos 5.1 are Anthropic’s first models to watermark text outputs.
fulafel
Data retention still sounds bad: "Claude Fable 5.1 and Claude Mythos 5.1 carry 30-day data retention and aren't available under zero data retention unless expressly authorized by Anthropic."
Anyone know who the ZDR special treatment is available to?
show comments
spondyl
Somewhat ironically, Fable 5.1 was flagged by the biology safeguards after I asked it to have a dig around the Fable 5.1 system card :)
show comments
olirex99
I suggest you to give a look to the MCP protocol for hardware that is being proposed by Anthropic. The hardware will be the next harness of LLMs, they will be able to operate machines to reinforce their theories.
I still think that a major problem is that biological processes are not “fast” as coding, but they are verifiable. If during post processing we are able to give enough harness to test and verify this kind of environment (maybe via simulation and real data) we will for sure achieve incredible performance also in this domain.
show comments
miki123211
> These patterns invalidate every later thinking block:
• [...] Rebuilding the top-level system prompt or tools array between requests in the same conversation.
Many people unknowingly do this (at a high cost to them because of the cache busts), this change will finally force them to stop.
Especially if you're generating your system prompt via a template that can change mid conversation, it's so easy to fall into this trap.
jebarker
Interesting that Fable produced the Venus elevation map by training a neural net to generate it. I wonder what the prompting looked like to make that happen, I.e. was this a spontaneous discovery or the result of a specific request.
spicypixel
Yeah but haiku 5 when?
show comments
Husafan
I find myself wondering how much of the writing style is based on financial incentives?
When paying by the token, don't the labs have a strong incentive to make the model as verbose as possible?
2001zhaozhao
There's now a 40X discount in the cache input pricing instead of 10X.
This seems to point to them having achieved some kind of optimization in attention mechanism perhaps along the lines of DeepSeek V4, which had a similarly high discount between cache input and normal input.
In real world use, the savings should be quite noticeable. For example, you can now use the model at 800K tokens context window at the same cost efficiency as the previous model at 200K tokens context window.
mohitpaddhariya
Interestingly, Claude’s output is now actually readable with Fable 5.1. Pretty sick.
exabrial
Did we get thought traces back? If no, it's useless.
show comments
vlovich123
> In part, this is because Fable 5.1 can now be used to discover software vulnerabilities—though not to develop exploits for them
Generally once an exploit chain is described, developing the exploit is trivial.
If you're so inclined, discover the exploits using Fable 5.1 and then give that exploit to a model that doesn't have such compunctions (e.g. local LLM or an uncensored cloud model / model that's easier to jailbreak). I don't think Anthropic is really mitigating here anything in the real world other than PR narratives where media can report "Anthropic's model was used to develop the latest cyber attack".
nirmeet011011
Great,excited to use these models
koolba
> Data retention. Our new system of Enterprise Frontier Safeguards (EFS) gives customers complete privacy (the same as a zero data retention policy) while still being state-of-the-art at preventing adversarial use. EFS works by storing data in cloud infrastructure controlled entirely by the customer, not Anthropic. It will be made available to enterprise customers in phases, beginning later this fall. Until EFS is available, eligible customers will be able to use Fable 5.1 with zero data retention.
This is interesting. I wonder if customers will be allowed to create an auto expiry for their own data to prevent future subpoenas. That’d be a treasure trove for discovery.
Bluestein
Unless these people start offering free, unlimited inference for a cautionary period so we can test the new model without an up-front (re-)investment, I am not touching this load-bearing pile of neuralese spew with a ten thousand token pole.-
show comments
leecommamichael
I'm having a very hard time finding mention of token-generation speed.
6thbit
Even with discounted cache, their prices remain way above everyone else but not necessarily the results.
What exactly is the premium that you're getting for paying these prices?
bix6
Why aren’t these models available on subscription plans?
I tried the old fable and it didn’t seem worth paying for. It still made errors like Opus does so I might as well use the included model…
show comments
niteshpant
I don't know how I feel when all the documentations are written by AI for humans.
AI to AI doc share: sure, do what you please.
AI to human: please make it legible and flowly.
example, "Every thinking block records which model produced it, and it's preserved in one direction only: Claude Fable 5.1 reads earlier models' thinking blocks, and no earlier model reads Claude Fable 5.1's." is a very Claude-isk way of writing. Choppy, long, and lacking flow.
5555watch
Is Fable 5.1 still actively downthrottling the reasoning when questions relate to frontier ML questions, like it did with 5.0?
ilia-a
Unfortunately at the moment the model is very quickly burning through plans, single session with 3-4 subagents, none using Max or Xhigh, mostly medium + some High can burn through 5h limit within 20-40 minutes of usage.
ckugblenu
This coupled with verification primitives will be quite compelling. we really have to start reimagining existing systems and processes from the ground up.
bilsbie
Will it still refuse my mitochondria questions?
cromka
"with cache reads at a quarter of the cost"
OK, I think that's what they meant when they suggested reduced extra promo usage will not sting this much.
mentalgear
> Forced tool use is not supported
That seems unfortunate for 3rd party integrations that expect stable output - what that really necessary ?
joduplessis
Anthropic, the company employing "treat them mean, keep them keen" as a marketing tactic. Pass.
finnjohnsen2
OpenCode+GLM-5.3 (and 5.2) for three weeks solid. Im so happy I made it out
nightsd01
I have to say, I am quite frustrated with Anthropic lately. I so badly want to use Fable to work on a side project of mine, which I used to do previously with no issues. But lately, they must have made some classifier change because it keeps hitting their stupid, overly-hyper-aggressive safeguard due to 'general_harms'.
Guys, listen to your feedback please. I hadn't used OpenAI products in quite a while until this issue came around. They seem to have MUCH smarter safeguards than Anthropic does.
Fordec
Going to hold off a few days until I adopt it, lets see what the general consensus develops as. Regretted jumping over day one for 5.0. The caching thing seems the most useful, but doesn't change anything for my subscription.
dmix
I use Claude Design heavily, I wish these charts show "10% better at picking a color" or laying out an app. Maybe it's hard to build a good visual design test. Claude's good at layouts but not the colors or smaller design details.
show comments
loeg
Are we getting a new Opus 5.1, then?
stillpointlab
My only concern is that sooner or later the best models will be priced out of my ability to pay.
I have been happy with Fable 5, it has done great work for me so far. Very excited to try out Fable 5.1 and see what differences and improvements there are.
show comments
TuxSH
Unfortunately isn't included in subscriptions and requires usage credits...
mixedbit
I'm afraid watermarking could restrict applications where LLMs can be safely used to assist with writing. If I write something myself and use an LLM to proofread it, without watermarking I can confidently say that corrections done by LLMs are small and insignificant enough to claim that the text is still authored by me, not by the model. With watermarking, however, I will never be sure if the result will not be flagged as AI generated, even if the AI contribution is very minor.
show comments
amluto
Looks like the API is nerfed to mitigate some recent thinking extraction attacks.
I wonder to what extent this will make the automatic Fable-to-Opus downgrade give worse results.
noduerme
I'm confused about Anthropic's pricing. Can anyone explain why Sonner 5 is $2/MTok in and Sonnet 4.6 is still $3?
show comments
ghoshbishakh
"Content provenance" seems to be activated with this model.
ayhanfuat
I noticed they reset the usage and I was kind of happy because this week it was using my quota much faster; I assumed they fixed that. Apparently it is for the celebration of 5.1?
show comments
cdnsteve
The average company and definitely average Joe will never be able to afford is ludicrously expensive model. Do not use this in a corporate/startup environment unless you have endless VC cash.
mrcwinn
If Anthropic thinks Opus 5 is very good, it is a window into how insular their culture is. I find it far, far behind Sol. It’s downright annoying to use.
genxy
This change was for them, not us. I am touching grass until next week while they get this shit sorted out. Not on my time.
Exoristos
Am I alone in not prioritizing the quality of prose produced by my coding agent? My foremost and almost only concern is how well it can engineer software.
show comments
leumon
Fable 5.1 seems to be the first model who can accurately draw an airbus a320 in 3D space given a set of limited tools (a brush with params color, size hardness and xyz coords): https://youtube.com/shorts/vyHsMqop2yw
jgilias
Cool. I’ve realized though that I don’t really need better models anymore. SOTA is good, I just want them faster/cheaper now.
ramon156
I yearn for a model that can churn through claude text and write sensible text. so far gemini is pretty good at that, even in the low variant
show comments
yoanwaidev
Almost finished my weekly limit today! I am more excited from the usage reset!
nubinetwork
Not until you stop being cheap and let pro users use fable under their existing paid subscriptions.
charcircuit
The safeguards and required extra retention is still not gone. Further more they are working to create separate tiers of access with the new biology program instead of giving everyone equal access to AI. Anthropic once again are showing they can not be trusted.
wewtyflakes
The breaking API changes are frustrating, especially the one that removes forced tool use.
george_max
"Price. Fable 5.1 will cost an estimated 25% less than Fable 5 for typical workloads, wherever usage is billed by token. This is because we’re reducing our pricing on cache reads (where the model reads inputs that have already been processed and stored). For highly agentic work, the savings will often be much larger—up to approximately 45%."
They show this off, but artificial analysis contradicts the statement. Fable 5 cost $3.14 per task, while 5.1 cost $3.69 -- around a 15% jump in pricing.
These, IMO, are marginal improvements for a more expensive model. I stopped using Claude ~3 months back; its outputs are too jargoned, it makes architectural decisions that are not right, and it's incredibly pricey for what it is. Each decision it makes, it acts as if a problem as major as world hunger has been solved. And the overly verbose code comments, strange commit descriptions, duplicate code, and slop it generates -- which I know is not specific to Fable -- is just too much for me.
I found the best is to use something like Deepseek V4 Flash -- with a fast TPS provider -- and work on the code myself. For agentic work with computer use, GLM 5.3 flash with Hermes Desktop works well.
andai
The most remarkable thing here is just how close Opus 5 is on most of these benchmarks.
show comments
spwa4
Strange that the system card carefully seems to avoid any benchmark where you can also find scores for GLM, Qwen. There's barely any overlap with GPT 5.6 benchmarks. Just these:
Model HLE w/tools GDPval-AA v2
Claude Fable 5.1 65.0 1853
GPT-5.6 Sol 64.5 ~1711-1730
GLM-5.3 62.5 1769
DeepSeek V4 Pro 60.0 1590
Kimi K3 59.8 1682
Qwen3.8-Max 56.2 1739
pmdr
Should've just named them both Guardrails 5.1 and be done with it.
tosh
> The watermark doesn't change the meaning, quality, or readability of the output
how?
show comments
joshfraser
the counterbalance to the AI doomers has always been the fact that everyone has equal access to AI. i hate this new world where Anthropic believe they should be the ones to decide who gets access to super intelligence and who doesn't.
dainiusse
Don't care unless it is priced in as other models.
thway15269037
Why would anyone use Antropic with these prices and full of bullshit safeguards, where chinese models rarely have any at all and massively cheaper? You can't even ask it to pentest auth code it itself has written.
brcmthrowaway
When is Astra launching?
sergiotapia
$50/M output is wild as hell - I haven't been using anthropics models for months now but who is paying for these tokens??? How can you justify spending that much money?
purpleidea
> Enterprise Frontier Safeguards (EFS)
Sounds like some serious nonsense. "Tell me you want the government to retain access to my data without saying it explicitly."
enraged_camel
Interesting that they seem to have gone all-in on science, and life sciences in particular. Improvements to coding performance seem marginal, although cost savings are very welcome.
Curious to see how Astra does.
abroszka33
Looks like agentic coding plateaued, and agentic scientific research is the new hype?
tclancy
> Denser prose in places.
Really? Interesting choice. Pretty much every CLAUDE.md file I have starts with something about Hemingway, terseness and treating every word you use like you're carving it on your own back, but different strokes for different folks. I suppose I haven't heard from anyone who enjoys how wordy Claude is because they aren't done writing their post yet.
Computer0
This seems like a welcome change: Claude Fable 5.1 also supports changing effort mid-conversation with a per-message output_config, which preserves the prompt cache.
kosolam
Unfortunately, I just canceled my max account.
Unfortunately, for them.
tamimio
I think what’s the industry is interested to see now isn’t “the best and latest super intelligent frontier model ever!!”, but rather the ability to run good enough models locally or better, on consumer or laptop grade specs. So I am not that impressed, plus haven’t used Claude for a while nor planning to, their models are useless with their “safe guards”.
eis
According to Artificial Analysis, 5.1 cost 56% MORE than 5, $8523 vs $5455. Yes cache cost is lower but it was MUCH more verbose: 140M vs 83M output tokens.
This directly contradicts what Anthropic is presenting here. Yes it scores higher but that's to be expected from a new release. It's the opposite of what OpenAI has been doing which was reducing costs, increasing efficiency.
Goes to show what a farce the supposed paradigm shift from Mythos and Fable was. All marketing, as always.
philipwhiuk
> Claude Fable 5.1 follows explicit tool instructions reliably.
Moving stuff out the API into prompt engineering is obviously less reliable but necessary for progression to 'actual intelligence'. Will be interesting to see if it really is solid.
BoorishBears
At least half the changes are just anti-distillation strategies...
show comments
canadiantim
Thank the heavens for quota resets
eis
I am not sure if Fable is worth it, at least with version 5 vs Opus 5. Opus beats Fable in quite a few benchmarks and at twice the cost I just haven't seen it provide noticeably better results compared to Opus. Has anyone noticed big differences? I did notice Opus maybe making more mistakes repeatedly but I don't have hard numbers on this. I hope Fable 5.1 brings noticeable improvements. I am giving it a go now on my 20x Max plan on a problem that Opus 5 has struggled for more than week now and has made very slow progress with regular regressions on the way.
show comments
danieltk76
The guardrails are horrendous for cybersecurity. you will get booted quickly down to Opus 4.8
dfltr
This feels kind of petty, but what is going on with those fuckass clouds in the background? Did no one notice how uncanny that whole thing looks?
thisisauserid
Zero data retention coming soon!
... with the condition that you store 100% of your data and make it available to the US government and possible others.
sashank_1509
Reads like AI slop, surprised they can’t see it in their blog post. No human wants to read in such prose
scronkfinkle
Has anyone been able to get anything substantial done with Fable in the first place? I more or less had totally given up on using it since the alignment checks were so sensitive that it pretty much always threw me back to Opus.
show comments
ike4est
weary of trying this model out after the amount of requests Fable 5 sent to Opus 5 which created for a terrible UX IMO.
lousken
While haiku is almost one year old. What a joke
show comments
tusimi
aaaaand its blocked from doing even basic tasks in biotech...
robinpie
do you think we'll go a full year without a new haiku lol
show comments
zb3
So bullshit safeguards are still there.
2001zhaozhao
Hi Claude, please cure aging, make no mistakes
show comments
YCisDead
dev.to > hacker news
For real devs, it’s way better.
HN is all Claude and Israel bullshit now
MadsRC
I was looking forward to using Fable for cybersecurity work, but kept getting bumped to Opus… Signed my org up for CVP, went through the trouble of procuring a separate team plan from our main org as Anthropic can only disable cyber safeguards for an entire org and not individual users…
After months of trouble dealing with KYC and procurement I finally got CVP for my security org and today I found out that CVP (which is what removes cyber safeguards) does not apply to Fable…
So yeah, unless you’re a Project Glasswing member, there’s no using Fable (which with Glasswing is Mythos) for security work… Absolutely useless…
Didn’t they just sign some “we must use AI for cyber defense before the bad guys do” and then they artificially cap us by not allowing Cyber-unlocked Fable…
Sigh…
show comments
krupan
Why is a marketing press release for a propietary product number 1 on hacker news. Again.
eigenblake
I am absolutely thrilled that they reset weekly limits. I have been experimenting with highly autonomous work (5+ hours continuous) and fable seems excellent at this, especially when using subagents. I ran out of Fable capacity and was bummed out that my experiment would take longer to complete. Now I'm super happy I get to continue it
(I work at Anthropic)
Beyond all the benchmarks, I think Fable 5.1 is a big improvement in writing style. It sounds a lot less stereotypically like other Claude models, has (imho) a much more natural style, and responds to my style instructions more reliably. More work to be done (and we will!) but reading better prose makes me so much happier.
Another point I expect not to get much attention until it all happens at once is science. People have been correctly excited about the many "sudden" breakthroughs LLMs are making in Maths, but some of the science benchmarks make me believe we'll soon see similar developments in other scientific domains. Fable 5.1 more than doubled Fable 5's Terminal-Bench-Science [1] score, which I think is meaningful.
[1] https://github.com/harbor-framework/terminal-bench-science
Pelicans for thinking effort low, medium, high and xhigh (that xhigh one is pretty good): https://tools.simonwillison.net/markdown-svg-renderer#url=ht...
I'm still waiting for effort max to finish.
EDIT: I fixed a bug in my tooling so it now records summarized reasoning traces - here's that max pelican, which is a significant improvement: https://tools.simonwillison.net/markdown-svg-renderer#url=ht...
Took just under 14 minutes to generate, and at 65927 output tokens cost me a hefty $3.30!
Excerpts from the reasoning trace:
> Adding pedal shapes near both feet, with the far foot on the second leg partially visible behind the frame. I'm considering whether to add a small scarf or cap for extra character, but leaning toward keeping it simple to avoid clutter.
> Now I'm debating a bicycle helmet on the head versus the pelican's signature crest—the beak and pouch already read clearly as "pelican," so a helmet could reinforce the bicycle theme without losing identity, though it might compete with the crest for visual space.
> I realize the beak at (484,84) would overlap with the dome helmet, so I need to shrink the helmet so it only covers the top of the head, adjusting its arc endpoints to sit higher and narrower so the beak can attach cleanly at the front without collision. [...]
> I'm adding a darker tip region to represent the primary feathers, then reconsidering the trailing edge to include scalloped feather curves instead of one smooth line for a more natural look. [...]
> Now I'm checking the vent line placements on the helmet, making sure they sit far enough inside the helmet's edge given the stroke width and rounded caps, and confirming each vent stays within the helmet's circular boundary. [...]
> I decide skipping a handlebar bell and tire highlights since they're unnecessary additions. Now I'm reconsidering the front fork's curve — the current control point pulls the shape backward when it should bow forward for a proper rake, so I need to shift the control point rightward to fix the fork's lean.
This is a notable result because most of the recent Claude models have been pretty bad at drawing pelicans, at least when compared to models in the Gemini or GLM series.
The price reduction comes from the cache read pricing falling from $1/M to $0.25/M, which means that Fable 5.1 now costs half of Opus's cache read costs ($0.5/M).
This gives a lot of credit to the theory that Anthropic did not get much bite on Fable at its original pricing, which in turn likely places a ceiling on LLM pricing in general.
Interestingly also, if you take away terminal-Bench-Science 0.1 results, it is hard to see ANY improvement:
Terminal-Bench 4.0: Fable 5.1 is +3.5% vs Opus 5.
GDPval-AA v2: +1.5% vs Opus 5.
OSWorld 2.0: +2.5% vs Opus 5.
Humanity's Last Exam (with tools): +1.6%
Keep in mind that this is supposed to be an entirely higher tier of a model than Opus 5. For one tier up and one version up, these are not really improvements. Probably leaves no room to place Opus 5.1 anywhere. Combined with the fact that they are selling 'readability'... Has frontier progress finally stalled?
Anyone ever seen the SouthPark episode making fun of Game of Thrones: A Song of Ass and Fire? Anthropic's announcements reminds me of "The Dragons Are Coming" running joke.
What they have done:
* Nerfed Fable, as many of noted it's useless
* Leverage Mythos as a marketing strategy, claiming its too good to release
* Removed thought traces, one of the only useful things to make sure your prompts are working correctly
* Continue tons of hype about how good they are without delivering, going to great lengths to publish how their model "hacked" its way out of a sandbox they misconfigured.
* Push a bunch of EU Overregulation onto the rest of the world with text watermarking, decreasing quality of answers
Last year, they were at least focused on making improvements. Nowadays its just a bunch of handwaving at the church of how good they are.
The only saving grace is Opus 4.6 is still available. Just sucks we haven't seen any measurable improvement, despite all of the ceremony.
I am finding that I am now less interested in better models than I am in token budgets. My issue with Anthropic models now is that I don't feel like I can rely on them as a daily driver because they'll dry up before my quota resets.
I am becoming dependent on AI to make a living, and I need predictable spend on it. If I know I can't use a model regularly all month, my enthusiasm is limited.
I urge Anthropic to get better at this aspect of their business so I can come back to it.
Looks like all three breaking changes are patches for inadvertent chain of thought disclosure. Someone found out (don't have the tweet handy) that if you created a bogus "think_deeply" tool and then forced the model to use it, it would output what is believed to be its raw thinking there - I believe the first breaking change stops this. The second two are aimed at people getting Haiku to repeat thinking blocks from other models verbatim (since it can see the decrypted version). I get that in their eyes it's an "exploit" but still kinda disappointing that they patched this
If I am reading this right, Fable 5 was worse than Opus 5 in almost every category, while consuming twice the tokens? The things you learn every day...
I let it go a few hours on a not trivial but well-known problem, and it felt like it was just a little too plodding and just kind of mucked around a little too much and wasn't aggressive enough about getting stuff done. I asked it to wind it down and finish up and it took another hour and 15 to actually stop and commit without really getting much more done. Not very impressed here, if you can't tell. This new version also seems like (maybe this is written somewhere, I don't care to look.) this cycles through compactions every ~250k tokens which, I guess, seems like it might save Anthropic money on KV cache but does net me anything be pretty frequent pauses. (It didn't seem to lose the thread, at least.)
Not gonna say I want 5.0 as an option still... but maybe I do.
> For example, in testing by the investment firm Millennium, Fable 5.1 found the cause of a rare crash on their internal systems that none of their engineers (or any other model) had been able to explain after several years of trying.
Say what you will about LLM-generated code, but stories like this give me hope that software will never be as buggy as it once was.
I’ll be very excited to try it out and see the actual improvement in writing style. The denser writing style probably won’t bother me.
Anthropic seems to be listening to community complaint on HN about how the writing style is grating. And apparently the solution from Anthropic is to add this block to every conversation!?
> Mannered prose substitutes metaphor and flourish for direct statement. Instead of "a parameter worth varying," the mannered writer produces "a dial worth turning." Instead of "this point still matters," they write "this point earns its keep." The phrases exist to display the writer, not to convey the idea, and readers can tell. That is why mannered prose irritates: it makes the reader work harder so the writer can perform. It is also imprecise. Metaphors drag in connotations the writer did not choose and cannot control. The fix is to say what you mean. When a literal phrase is available, use it.
The above was quoted verbatim from https://platform.claude.com/docs/en/build-with-claude/prompt...
“ Claude Fable 5.1's writing is generally a step up from earlier Claude models, with fewer stock phrases and less unexplained jargon. In some cases, though, its prose is denser than Claude Fable 5's: sentences run longer and there are fewer paragraph breaks.”
I cancelled my pro max Claude subscription last week; codex is much more succinct. I am curious if this is getting better.
I don’t think Anthropic realizes that humans have a token limit too and it can be exhausting to read Claude’s output. Prose density is not the same thing as succinctness.
Instead of a new model that's going to have unreasonably shallow usage limits, I wish they would:
1) address the claude 20x plan usage being only 6-7x the ceiling of the claude pro plan
2) either fix opus 5, make it completely free, or delete it entirely
AI is really not "just software" anymore. It is able to discover facts and advance science. Hard to disagree that we're near or at the point where Artificial Intelligence has expanded reality into 4 quadrants:
objects that are not alive: dust, rocks, water, wood, hats, lego, aluminum, etc.
objects that are alive but not intelligent: trees, mold, staphylococcus, cancer, grapes, etc.
objects that are alive and intelligent: cats, Steven Tyler, dolphins, crows, dogs, elephants, etc.
and now intelligent but not alive: Fable, Grok, GPT, etc.
To be honest, these frontier model releases have become boring for me. Opus 4.8 was already good enough for most of my use cases. I don't have any projects right now that I would use Fable for instead of Opus. So when I see announcements like this I just think "that's cool I guess" and then go back to using weaker/cheaper models.
What's far more exciting right now is models like DeepSeek V4 Flash and GLM 5.3 Flash. They have achieved good-enough-intelligence at extremely low prices and fast speeds. I don't have a use for Fable-level intelligence, but I do have uses for Opus-4.8-level intelligence that I can use as much as I want without worrying about the bill.
I've been building Cargo-for-C (https://github.com/tspader/spn), and the difference between Fable and Opus was already astounding. Fable was the first time that I could point a model at a piece of code I'd written and expect it to make it meaningfully better rather than a hard pattern match to whatever mistakes it had.
5.1 so far seems like another leap, which is really surprising. I threw it at a few bigger features I've been designing for a while, and it came back with some extremely thoughtful wrinkles in the design that I'd legitimately not considered. Which, OK, package managers and build executors and compiling C/C++ is pretty well trodden ground, but my thing is very different from everything that exists, and I was very surprised it was able to understand all that context so deeply and intuitively
All the benchmarks in the world don't matter if the model just straight up refuses to do mundane things. Claude has too much of an attitude.
Fable 5.1 is actually more expensive than 5.0 when run on the Artificial Analysis suite:
https://artificialanalysis.ai/#intelligence-efficiency-tabs
From the changelog:
Whole-file rewrites for small changes. When editing text files, the model is more likely to rewrite the entire file than make a targeted edit. The result is usually the same, but the rewrite costs more output tokens and time.
So we are to catch that somehow? And then add their recommendation (below) to our prompts?
https://platform.claude.com/docs/en/build-with-claude/prompt...
If Claude Fable 5.1 rewrites whole files for small changes, append the following instruction to the system prompt or the first user message. Claude Fable 5.1 is more likely than Claude Fable 5 to rewrite an entire text file rather than make a targeted edit. The resulting file is usually the same, but unless the file is short or most of it is changing, a rewrite costs more output tokens and time. The instruction brings Claude Fable 5.1 back in line with Claude Fable 5 for small and medium changes.
> The number of tokens used to edit files is best minimized, all else being equal. Therefore, when it will not affect the end result, try to surgically edit a file rather than rewrite the entire thing.
> We’re introducing Claude Fable 5.1 and Claude Mythos 5.1. They’re the world’s most advanced models for coding and knowledge work—and their research capabilities offer an early glimpse of how AI models will contribute to scientific progress.
I'm not an emdash hater but this isn't how you use them. It should be a comma.
I've recently been running these agent sessions on more and more long running tasks because these latest models can do a REALLY good job on big chunks of work, and i've been watching them way less. It's starting to occur to me the importance of alignment is a today problem, it's not a tomorrow problem.
In the past I watched and saw everything the model did, not a lot got past me. Today it does A TON of work while i'm busy on other tasks. It also has extensive access to my computer, other computers on my network, my internet. It's really helpful when you give it a lot of resources, but right now I have very autonomous, very smart agent running around more or less unattended with a lot of resources.
What I don't see in the comments: "I had a specific problem I couldn't solve with the previous version of this LLM. But the improvements in this version unlocked the solution for me."
What I do see in the comments: subjective improvement in text generation, possibly lower cost, some optimism about code generation, but some skepticism too.
I use coding agents. To me they are very useful. But what I spend on them isn't going to support trillions of dollars in investment.
> This required us to add a watermark—a numerical way of determining the likelihood that Claude was involved in writing a piece of text—to the outputs of models released after August 2, 2026. As we recently explained, this watermark is invisible to anyone who does not have the detection API. It has no practical impact on the quality or content of Claude’s outputs and contains no information about the user, their organization, or their conversations with Claude.
How does this work if it doesn’t change the output?
Since one of the big improvements here is supposedly the writing style, on that topic I'm mystified about something:
Why is it that the voice models in Claude and ChatGPT have a perfectly normal style with barely any "AI smell", while the writing models are so obviously recognizable as AI?
The answer is likely that models underlying the voice modes are (post) trained differently. If so, then why can't the writing model be similarly trained? Presumably they haven't found a way to train them to be both "smart" (i.e. solve tasks etc) and pleasant to talk to?
Just don't expect to do any work on hardware/firmware you own with fable, I can hardly even type in the word "firmware" without it downgrading to Opus 4.8, which is totally unsatisfying. This even happens with Opus 5. Definitely making multiple classes of users moving forward and most of us are obviously going to be part of the permanent underclass.
Interesting that the "frontier" keeps moving forward from the guys who want a pause.
"Claude Mythos 5.1 is identical to Fable 5.1, but it offers more permissive safeguards for vetted individuals and organizations"
Then why does it have separate datapoints for Terminal Bench, and score higher? Something doesn't add up here??
"Cache reads now cost 75% less, or $0.25 per million tokens." For me, at a typical 95% cache hit rate, I think my optimal context window size before autocompaction goes from ~200K to ~400K tokens. Great for longer horizon tasks.
"In biology, we’ve established an access program, developed in partnership with the US government, to enable access to Claude Mythos 5.1’s advanced biology capabilities"
Having lived through Covid, this doesn't sound so good to me.
"Distillation is a safety risk, since the distilled capabilities can subsequently be released without adequate safeguards."
Can't believe they haven't at least figured out better messaging. If we take them at their word, it's hard not to read it as a messiah complex, that they think they're the only ones capable or worthy of making these decisions. I don't believe them, but I wouldn't be surprised if the articulated reason is a version of "distillation is a safety risk because we might lose the race".
Plus, completely deaf to the recent OpenAI-HF hack incident. Recall, defenders were categorically unable to use western frontier models in their response.
I was originally going to complain about the chem and bio guards still being too onerous, but I'll admit the projects Fable 5 categorically refused to work on are now usable, at least not rejecting on first prompt because the word "virology" was in a git commit (absolutely serious, in one repo it triggered on literally any prompt, eventually traced to the system prompt loading git commit history). Still, them trying to get into the biomed business while walling off the capabilities to the public reeks. Why sell the segments that are actually valuable if you can capture the value yourself!
My main gripe with LLMs is the cringe AI phrasings that they use in UI elements. Pompous things like "Your keys, supercharged" or weird yoda-speak stuff like "searches the app remembers" instead of just naming the thing "Learned searches".. you know, proper GUI copy like it was done for the past decades.
I jumped when I saw a mention about "writing style improvements" so I gave it a try on a recent feature in rcmd [0]. I prompted Fable 5.1 to find these wordings and propose simpler plain language.
It took every string including the ones I already rewrote by hand, and proposed even more weird LLM speak. Like for "Left Command conflict detected" it proposed "This keyboard can't tell left from right".It's a very capable coding agent, but I can't understand how it can be so bad at writing. Where are all these verbal tics coming from and why is it so hard to get rid of them?
[0] https://lowtechguys.com/rcmd
I'm so suspicious of this after Opus 5 benchmarks scored it higher than Fable 5, yet Opus 5 was untrustworthy (overconfident, error-prone).
Fable is too expensive for general use I think this is why it hasn’t received as much attention as expected since Fable came out Developers always work while trying to find ways to work continuously for a 5-hour session without disconnecting. Fable has had the experience of using up all its tokens before I even realized it because the burn rate was too fast. Since then, I always use only Opus. For Fable to become a common coding environment, it will have to reduce token consumption significantly compared to now
On both my work (Team Premium) and personal accounts (Max 20x), Fable 5.1 hit the 5-hour limit before it could finish the first task I gave it. On my work account, it took about 30 minutes, and on my personal account, less than an hour.
This has never happened to me before, but if this is normal behavior, Fable 5.1 is essentially unusable.
The thing with Fable-level models is that I will never feel comfortable using them for agentic tasks on a pay-as-you-go API pricing plan without monitoring them strictly, which becomes a chore.
I once caught Fable 5 spinning its wheels on a rendering issue, which evaporated 90% of my usage in a single prompt. I could never let Fable run free attached to a credit card without staring at it the whole time.
Bit of a discount if you're using caching:
> same input and output prices, with cache reads at a quarter of the cost
This should impact any long-running agent since subsequent calls can benefit from cached reads for previous transcripts.
Let me guess: it's the end of the world again. These new models are sooo powerful that will take over the world, just like the others before them.
Are they going to try the banned for export for a week marketing move too?
I cancelled my pro max 20x subscription, tired of Opus stopping the work from time to time, or saying "this is 2 months of work"
I wish I could afford Fable.
I am using Claude and Claude code for my own amateur history project. I'm enjoying how it constantly reaches dead ends, and I can reframe the question and get more results. I am starting to get concerned that AI and me are so compatible, that I might not be a human at all...
I also like that, because I'm too lazy to write stuff up, Claude code can keep the current state of research published on my site. It makes running a hobby site a dream. "I just found these pictures. Add them to the site for me". And up they go, resized and all. What a dream of a way to work. "Some of links in this article are dead, run through them and check, and see if you can get an archive link for me if they don't". It's like sending a Teams message to my PA.... which I don't have in real life
Sadly still not available for Pro subscription. At least they reset everyone's limits.
According to the FrontierCode Extended benchmarks in the system "card" (page 169-170), Fable 5.1 apparently does best on the medium effort level for this benchmark: "[...] at higher efforts, Fable 5.1 occasionally adds more small, unrequested changes [...]" Though Fable 5.1's medium is also lower than Fable 5's best score on the same benchmark, which uses xhigh.
I’m really excited to try this out. Fable and Opus 5 constantly wow me when working together. Unfortunately, I’m a little burned because of technical issues.
Anthropic accidentally over-billed my account, and when I reached out to the support bot, it downgraded my account to a Free account. It’s been impossible to get it resolved and I have almost $200 held hostage.
I don’t want to do a charge back. I’m one of the main advocates for Claude Code at work, I use this subscription to try out new features before it’s available at work.
The whole experience has been illuminating about our dependencies on these AI companies.
> On complex asynchronous workloads, though, nudge it not to end its turn before the work is done. Without the nudge, the model sometimes describes what it would do next instead of doing it ("Next, I'll …") or stops to ask permission for a step the original request already covered ("Shall I apply this?"). [1]
Interesting behavior. The docs also provided recommended prompt [1] to mitigate this behavior if undesired.
Wondering if anyone has encountered it yet?
[1]: https://platform.claude.com/docs/en/build-with-claude/prompt...
I think the most interesting thing about this, that I can tell so far, is the cache hit discount. Anyone who had an autocompaction threshold optimised for their use case should consider upping it from where it is.
I would be interested in whether someone has done research here on these things as it seems a fairly complicated function to work out, and use case dependent. (?)
In some sense an expired kv cache is basically like an expensive cache hit, so your compaction token threshold should come in. Ideally claude code should allow you to vary the autocompaction threshold to vary with time since last token, but it doesn't of course. This perhaps suggests that someone should manage claude code through their own intermediary agent who manages these sorts of rules.
Lastly, I strongly suspect that anthropic isn't offering this price cut out of the kindness of their hearts. I am sure that they are to some extent banking on people not reacting to their price cut and leaving their autocompaction thresholds unchanged.
[edit - looks like the discount is only for the api, so they still don't give a rats ass about subs!]
Worth noting: Claude Fable 5.1 and Mythos 5.1 are Anthropic’s first models to watermark text outputs.
Data retention still sounds bad: "Claude Fable 5.1 and Claude Mythos 5.1 carry 30-day data retention and aren't available under zero data retention unless expressly authorized by Anthropic."
Anyone know who the ZDR special treatment is available to?
Somewhat ironically, Fable 5.1 was flagged by the biology safeguards after I asked it to have a dig around the Fable 5.1 system card :)
I suggest you to give a look to the MCP protocol for hardware that is being proposed by Anthropic. The hardware will be the next harness of LLMs, they will be able to operate machines to reinforce their theories.
I still think that a major problem is that biological processes are not “fast” as coding, but they are verifiable. If during post processing we are able to give enough harness to test and verify this kind of environment (maybe via simulation and real data) we will for sure achieve incredible performance also in this domain.
> These patterns invalidate every later thinking block:
• [...] Rebuilding the top-level system prompt or tools array between requests in the same conversation.
Many people unknowingly do this (at a high cost to them because of the cache busts), this change will finally force them to stop.
Especially if you're generating your system prompt via a template that can change mid conversation, it's so easy to fall into this trap.
Interesting that Fable produced the Venus elevation map by training a neural net to generate it. I wonder what the prompting looked like to make that happen, I.e. was this a spontaneous discovery or the result of a specific request.
Yeah but haiku 5 when?
I find myself wondering how much of the writing style is based on financial incentives?
When paying by the token, don't the labs have a strong incentive to make the model as verbose as possible?
There's now a 40X discount in the cache input pricing instead of 10X.
This seems to point to them having achieved some kind of optimization in attention mechanism perhaps along the lines of DeepSeek V4, which had a similarly high discount between cache input and normal input.
In real world use, the savings should be quite noticeable. For example, you can now use the model at 800K tokens context window at the same cost efficiency as the previous model at 200K tokens context window.
Interestingly, Claude’s output is now actually readable with Fable 5.1. Pretty sick.
Did we get thought traces back? If no, it's useless.
> In part, this is because Fable 5.1 can now be used to discover software vulnerabilities—though not to develop exploits for them
Generally once an exploit chain is described, developing the exploit is trivial.
If you're so inclined, discover the exploits using Fable 5.1 and then give that exploit to a model that doesn't have such compunctions (e.g. local LLM or an uncensored cloud model / model that's easier to jailbreak). I don't think Anthropic is really mitigating here anything in the real world other than PR narratives where media can report "Anthropic's model was used to develop the latest cyber attack".
Great,excited to use these models
> Data retention. Our new system of Enterprise Frontier Safeguards (EFS) gives customers complete privacy (the same as a zero data retention policy) while still being state-of-the-art at preventing adversarial use. EFS works by storing data in cloud infrastructure controlled entirely by the customer, not Anthropic. It will be made available to enterprise customers in phases, beginning later this fall. Until EFS is available, eligible customers will be able to use Fable 5.1 with zero data retention.
This is interesting. I wonder if customers will be allowed to create an auto expiry for their own data to prevent future subpoenas. That’d be a treasure trove for discovery.
Unless these people start offering free, unlimited inference for a cautionary period so we can test the new model without an up-front (re-)investment, I am not touching this load-bearing pile of neuralese spew with a ten thousand token pole.-
I'm having a very hard time finding mention of token-generation speed.
Even with discounted cache, their prices remain way above everyone else but not necessarily the results.
What exactly is the premium that you're getting for paying these prices?
Why aren’t these models available on subscription plans?
I tried the old fable and it didn’t seem worth paying for. It still made errors like Opus does so I might as well use the included model…
I don't know how I feel when all the documentations are written by AI for humans.
AI to AI doc share: sure, do what you please.
AI to human: please make it legible and flowly.
example, "Every thinking block records which model produced it, and it's preserved in one direction only: Claude Fable 5.1 reads earlier models' thinking blocks, and no earlier model reads Claude Fable 5.1's." is a very Claude-isk way of writing. Choppy, long, and lacking flow.
Is Fable 5.1 still actively downthrottling the reasoning when questions relate to frontier ML questions, like it did with 5.0?
Unfortunately at the moment the model is very quickly burning through plans, single session with 3-4 subagents, none using Max or Xhigh, mostly medium + some High can burn through 5h limit within 20-40 minutes of usage.
This coupled with verification primitives will be quite compelling. we really have to start reimagining existing systems and processes from the ground up.
Will it still refuse my mitochondria questions?
"with cache reads at a quarter of the cost"
OK, I think that's what they meant when they suggested reduced extra promo usage will not sting this much.
> Forced tool use is not supported
That seems unfortunate for 3rd party integrations that expect stable output - what that really necessary ?
Anthropic, the company employing "treat them mean, keep them keen" as a marketing tactic. Pass.
OpenCode+GLM-5.3 (and 5.2) for three weeks solid. Im so happy I made it out
I have to say, I am quite frustrated with Anthropic lately. I so badly want to use Fable to work on a side project of mine, which I used to do previously with no issues. But lately, they must have made some classifier change because it keeps hitting their stupid, overly-hyper-aggressive safeguard due to 'general_harms'.
Guys, listen to your feedback please. I hadn't used OpenAI products in quite a while until this issue came around. They seem to have MUCH smarter safeguards than Anthropic does.
Going to hold off a few days until I adopt it, lets see what the general consensus develops as. Regretted jumping over day one for 5.0. The caching thing seems the most useful, but doesn't change anything for my subscription.
I use Claude Design heavily, I wish these charts show "10% better at picking a color" or laying out an app. Maybe it's hard to build a good visual design test. Claude's good at layouts but not the colors or smaller design details.
Are we getting a new Opus 5.1, then?
My only concern is that sooner or later the best models will be priced out of my ability to pay.
I have been happy with Fable 5, it has done great work for me so far. Very excited to try out Fable 5.1 and see what differences and improvements there are.
Unfortunately isn't included in subscriptions and requires usage credits...
I'm afraid watermarking could restrict applications where LLMs can be safely used to assist with writing. If I write something myself and use an LLM to proofread it, without watermarking I can confidently say that corrections done by LLMs are small and insignificant enough to claim that the text is still authored by me, not by the model. With watermarking, however, I will never be sure if the result will not be flagged as AI generated, even if the AI contribution is very minor.
Looks like the API is nerfed to mitigate some recent thinking extraction attacks.
I wonder to what extent this will make the automatic Fable-to-Opus downgrade give worse results.
I'm confused about Anthropic's pricing. Can anyone explain why Sonner 5 is $2/MTok in and Sonnet 4.6 is still $3?
"Content provenance" seems to be activated with this model.
I noticed they reset the usage and I was kind of happy because this week it was using my quota much faster; I assumed they fixed that. Apparently it is for the celebration of 5.1?
The average company and definitely average Joe will never be able to afford is ludicrously expensive model. Do not use this in a corporate/startup environment unless you have endless VC cash.
If Anthropic thinks Opus 5 is very good, it is a window into how insular their culture is. I find it far, far behind Sol. It’s downright annoying to use.
This change was for them, not us. I am touching grass until next week while they get this shit sorted out. Not on my time.
Am I alone in not prioritizing the quality of prose produced by my coding agent? My foremost and almost only concern is how well it can engineer software.
Fable 5.1 seems to be the first model who can accurately draw an airbus a320 in 3D space given a set of limited tools (a brush with params color, size hardness and xyz coords): https://youtube.com/shorts/vyHsMqop2yw
Cool. I’ve realized though that I don’t really need better models anymore. SOTA is good, I just want them faster/cheaper now.
I yearn for a model that can churn through claude text and write sensible text. so far gemini is pretty good at that, even in the low variant
Almost finished my weekly limit today! I am more excited from the usage reset!
Not until you stop being cheap and let pro users use fable under their existing paid subscriptions.
The safeguards and required extra retention is still not gone. Further more they are working to create separate tiers of access with the new biology program instead of giving everyone equal access to AI. Anthropic once again are showing they can not be trusted.
The breaking API changes are frustrating, especially the one that removes forced tool use.
"Price. Fable 5.1 will cost an estimated 25% less than Fable 5 for typical workloads, wherever usage is billed by token. This is because we’re reducing our pricing on cache reads (where the model reads inputs that have already been processed and stored). For highly agentic work, the savings will often be much larger—up to approximately 45%."
They show this off, but artificial analysis contradicts the statement. Fable 5 cost $3.14 per task, while 5.1 cost $3.69 -- around a 15% jump in pricing.
https://artificialanalysis.ai/
These, IMO, are marginal improvements for a more expensive model. I stopped using Claude ~3 months back; its outputs are too jargoned, it makes architectural decisions that are not right, and it's incredibly pricey for what it is. Each decision it makes, it acts as if a problem as major as world hunger has been solved. And the overly verbose code comments, strange commit descriptions, duplicate code, and slop it generates -- which I know is not specific to Fable -- is just too much for me.
I found the best is to use something like Deepseek V4 Flash -- with a fast TPS provider -- and work on the code myself. For agentic work with computer use, GLM 5.3 flash with Hermes Desktop works well.
The most remarkable thing here is just how close Opus 5 is on most of these benchmarks.
Strange that the system card carefully seems to avoid any benchmark where you can also find scores for GLM, Qwen. There's barely any overlap with GPT 5.6 benchmarks. Just these:
Should've just named them both Guardrails 5.1 and be done with it.
> The watermark doesn't change the meaning, quality, or readability of the output
how?
the counterbalance to the AI doomers has always been the fact that everyone has equal access to AI. i hate this new world where Anthropic believe they should be the ones to decide who gets access to super intelligence and who doesn't.
Don't care unless it is priced in as other models.
Why would anyone use Antropic with these prices and full of bullshit safeguards, where chinese models rarely have any at all and massively cheaper? You can't even ask it to pentest auth code it itself has written.
When is Astra launching?
$50/M output is wild as hell - I haven't been using anthropics models for months now but who is paying for these tokens??? How can you justify spending that much money?
> Enterprise Frontier Safeguards (EFS)
Sounds like some serious nonsense. "Tell me you want the government to retain access to my data without saying it explicitly."
Interesting that they seem to have gone all-in on science, and life sciences in particular. Improvements to coding performance seem marginal, although cost savings are very welcome.
Curious to see how Astra does.
Looks like agentic coding plateaued, and agentic scientific research is the new hype?
> Denser prose in places.
Really? Interesting choice. Pretty much every CLAUDE.md file I have starts with something about Hemingway, terseness and treating every word you use like you're carving it on your own back, but different strokes for different folks. I suppose I haven't heard from anyone who enjoys how wordy Claude is because they aren't done writing their post yet.
This seems like a welcome change: Claude Fable 5.1 also supports changing effort mid-conversation with a per-message output_config, which preserves the prompt cache.
Unfortunately, I just canceled my max account.
Unfortunately, for them.
I think what’s the industry is interested to see now isn’t “the best and latest super intelligent frontier model ever!!”, but rather the ability to run good enough models locally or better, on consumer or laptop grade specs. So I am not that impressed, plus haven’t used Claude for a while nor planning to, their models are useless with their “safe guards”.
According to Artificial Analysis, 5.1 cost 56% MORE than 5, $8523 vs $5455. Yes cache cost is lower but it was MUCH more verbose: 140M vs 83M output tokens.
This directly contradicts what Anthropic is presenting here. Yes it scores higher but that's to be expected from a new release. It's the opposite of what OpenAI has been doing which was reducing costs, increasing efficiency.
Fable 5: https://artificialanalysis.ai/models/claude-fable-5 Fable 5.1: https://artificialanalysis.ai/models/claude-fable-5-1
> Hey Cl… Your limit has been reached.
Goes to show what a farce the supposed paradigm shift from Mythos and Fable was. All marketing, as always.
> Claude Fable 5.1 follows explicit tool instructions reliably.
Moving stuff out the API into prompt engineering is obviously less reliable but necessary for progression to 'actual intelligence'. Will be interesting to see if it really is solid.
At least half the changes are just anti-distillation strategies...
Thank the heavens for quota resets
I am not sure if Fable is worth it, at least with version 5 vs Opus 5. Opus beats Fable in quite a few benchmarks and at twice the cost I just haven't seen it provide noticeably better results compared to Opus. Has anyone noticed big differences? I did notice Opus maybe making more mistakes repeatedly but I don't have hard numbers on this. I hope Fable 5.1 brings noticeable improvements. I am giving it a go now on my 20x Max plan on a problem that Opus 5 has struggled for more than week now and has made very slow progress with regular regressions on the way.
The guardrails are horrendous for cybersecurity. you will get booted quickly down to Opus 4.8
This feels kind of petty, but what is going on with those fuckass clouds in the background? Did no one notice how uncanny that whole thing looks?
Zero data retention coming soon!
... with the condition that you store 100% of your data and make it available to the US government and possible others.
Reads like AI slop, surprised they can’t see it in their blog post. No human wants to read in such prose
Has anyone been able to get anything substantial done with Fable in the first place? I more or less had totally given up on using it since the alignment checks were so sensitive that it pretty much always threw me back to Opus.
weary of trying this model out after the amount of requests Fable 5 sent to Opus 5 which created for a terrible UX IMO.
While haiku is almost one year old. What a joke
aaaaand its blocked from doing even basic tasks in biotech...
do you think we'll go a full year without a new haiku lol
So bullshit safeguards are still there.
Hi Claude, please cure aging, make no mistakes
dev.to > hacker news
For real devs, it’s way better.
HN is all Claude and Israel bullshit now
I was looking forward to using Fable for cybersecurity work, but kept getting bumped to Opus… Signed my org up for CVP, went through the trouble of procuring a separate team plan from our main org as Anthropic can only disable cyber safeguards for an entire org and not individual users…
After months of trouble dealing with KYC and procurement I finally got CVP for my security org and today I found out that CVP (which is what removes cyber safeguards) does not apply to Fable…
So yeah, unless you’re a Project Glasswing member, there’s no using Fable (which with Glasswing is Mythos) for security work… Absolutely useless…
Didn’t they just sign some “we must use AI for cyber defense before the bad guys do” and then they artificially cap us by not allowing Cyber-unlocked Fable…
Sigh…
Why is a marketing press release for a propietary product number 1 on hacker news. Again.
I am absolutely thrilled that they reset weekly limits. I have been experimenting with highly autonomous work (5+ hours continuous) and fable seems excellent at this, especially when using subagents. I ran out of Fable capacity and was bummed out that my experiment would take longer to complete. Now I'm super happy I get to continue it