Claude Opus 5

So the rumors were right, Opus 5 was indeed being polished up for release. Huge improvements in GDPval-AA v2 too -- great for some of the knowledge work-based agentic workloads I run.

Also glad they still kepy Fable 5 on "credits only" access. I think we're going to start seeing model providers gate top-of-the-line models behind pay-as-you-go API rates/credits while subsidizing other models on monthly subscriptions.

Also the cost per task. It appears to be significantly cheaper, cheaper than sonnet!

I don't understand how the data retention works. My company has an enterprise license with no data retention but if I ask Claude about past conversations, it remembers. So surely the information is being stored somewhere

> Updated over 2 weeks ago

I hope we get clarification on this, I can't find anything claiming that it is compatible with ZDR.

do you mean, that organizations now have access to Fable-ish pelican drawing?

insane pricing:

" Claude Opus 5 is available today on all platforms, priced at $5 per million input tokens and $25 per million output tokens (the same as Opus 4.8)"

[dead]

I like how they highlighted Opus 5 as the best for “Agentic Coding” even though the number is slightly lower than Fable. Close enough for marketing, I guess!

Thanks for that, looks really good. I can see why they were constantly pushing back fable going out of the max sub with these benchmarks

I wonder why FrontierCodev1.1's data lists Opus 5 as better than Fable 5.

You're comparing the "score percentage" (e.g. out of the total number of partial points available, how many did the agent achieve) to the "completion percentage" (how many tasks does the model score 100% on). The paper says "Claude Opus 4.8 with maximum thinking and batched tool calls scores best but still completes only 20.6% of tasks at a 54.8% partial score", which is ~the same number that Anthropic reports here (55.7 vs 54.8).

That is—the agent scored 100% on 20% of tasks, but on average it got 54% of the "score" awarded in the exam. One number reflects partial progress, the other one doesn't. The authors of the benchmark prefer you to look at the lower number (because they want to show their benchmark as capturing useful gaps in capabilities and with a lot of room for improvement), the authors of the models want you to look at the higher number (because they want you to think of their models as capable)

[flagged]

Its slop all the way down.

I think some variance is to be expected since LLMs are typically non-deterministic, however that's a huge difference that I think warrants further explanation.

Model Routing will always be done better by models themselves. Plus routing loses context making it more expensive and less reliable.

Model Routing is just Bitter lesson. The models themselves will get better at this and frontier companies will simply give that capability

Because they're trying very hard not to understand it.

Otherwise the expensive-yet-powerful model probably won't see much revenue. How much money is there in bleeding edge scientific research? There's a lot, but there's even more existing capital in paying people people to do college level paperwork, and the bulk of those traffic gets routed to the cheapest model.

You mostly don't need super powerful AGI to replace the paper pushers, but the frontier labs are trying to position themselves as being uniquely capable of producing super powerful AGI, and also be the ones replacing office workers.

Not sure how it will work out for them, but I think model routing is going to poke holes in that narrative. That's why I think they're trying very hard not to understand model routing exists.

Who are the customers though? Honest question, I'd like to understand it.

For me, anything other than current best available SOTA for any task is unacceptable. The only routing rule I need is "the most powerful model I still have flat-priced quota available for". I mean, why settle for less?

what's the threshold for model routing where you're willing to trust the router?

For coding my own work I don't trust the model router, and it would have to be shown to be to save a real dollar amount.

From a buying perspective it's a hard sell to save x but lose out on bugs you are probably introducing at an unquantifiable severity and frequency. How much is it worth to hedge your bets by doing every single inference request on the frontier model?

How much will it cost to go back later and fix things, but also the meta question of how to be able to decide on a hypothetical unknowable? (You'll never know how much better or worse your code was gonna be, it's untestable at a project level)

Openrouter should ideally kill in this space and make their model agnostic infra like memory, harnesses, chat applications.

Easy enough to explain: they're benchmaxxing. Fable is intelligent but not benchmaxxed. Opus is less intelligent but benchmaxxed.

Capable in term of AI R&D, not capable in terms of hacking (which caused all the Fable drama.) But agree, confusing wording.

Maybe they don't want to say that to avoid the government scrutiny.

Go read the safeguards section in the report and you will realize why that is.

These models are heavily as safeguarded and that was the initial reason why they said they couldn't and haven't released Mythos because that model is the one without the safeguards.

OpenAI is did the same thing when they announced a model without safeguards broken into HuggingFace servers.

Fable established the frontier, this is just catching up.

I'm not sure about the answer here, but this can be caused by the scoring rubric used by given benchmarks. For instance, if a benchmark docks scores for running too many commands or using too much wall-clock time, higher efforts will get lower scores.

It is probably from randomness. The benchmark tasks nowadays are so long that you can't really afford to run a large number of samples of them per model & effort combination

I agree, very odd they did not comment on any theories for the degradation here. Dip and then rebound at max effort is pretty interesting too. Overthinking is bad, but you can overthink so much it starts to be better again?

apparently it got docked points for editing files out of scope

On a small, easily digestible task, I compared Fable to Opus and the cost of Fable was easily 2x despite being fewer tokens, and the output was not really better. Obviously, there are tasks where using Fable matters but honestly they're rather unusual. And for a lot of tasks I've found downgrading to Sonnet can be valuable because Fable and Opus are a lot more secretive about what they're doing, and it's impossible to "listen to them think" and stop them when they start making off-the-wall inferences/assumptions and going down bad paths.

I think in the long run tokens are probably the wrong thing; it's compute and cache memory that you need to be measuring, and when you look at it that way I suspect in most cases the models have pretty similar performance.

This is a comment about user-facing responses, which are seldom the thing you're worried about when thinking about token efficiency.

> especially with OpenAI making so much progress with the efficiency of their models

To be fair though, Sol tends to go off the rails sometimes. It's much less reliable than Fable in its outputs. It tends to be overzealous in its research/changes.

> This still feels like a step in the wrong direction, though, especially with OpenAI making so much progress with the efficiency of their models

Gemini also had modest increase before this - don't be surprised when OpenAI also has a "modest increase" with its next release. Cartel-like behaviour doesn't require direct communication when none of the participants are interested in participating in a margin-destroying price-war. All one needs to do is raise their price and watch how the competition react.

Such a scheme (and resulting high margins) would be imperilled by the existence of frontier open-weight models in the market, which may be why the reaction to Chinese models may be particularly shrill.

This is part of why I switched to Grok 4.5

I don't need more powerful models, I need one that responds fast enough that my attention doesn't wander to other tasks. Grok 4.5 is so fast I can just use it in-band without swapping to other tasks.

Slower than Opus 4.8, which was already miserably slow, is indeed a step in the wrong direction.

Is that true? Sol responses are also longer than prior models.

The final user-facing responses are usually a tiny fraction of the total tokens used over the course of a given conversation turn. When you're doing any real work, reasoning and tool uses constitute the overwhelming majority of the tokens in / out... not the final user-facing response.

It's a step in the wrong direction but also token efficiency has become a focus relatively recently (just the past few weeks it seems like the zeitgeist has turned it's attention to efficiency) while work on this model probably started many many months ago. I would expect to see models released that focus on token efficiency in 6-12 months

Funny that a company selling an AI software developer can't use it to fix their infra.

Fixing those issues still requires humans.

> Constant bugs, dropped sessions, issues switching models, http errors. It's becoming ridiculous

And memory leaks.

> coding is largely solved

- Boris

I found the same thing funny with Computer Use from OpenAI. It struggled to open and close Spotify.

So dangerous! I can't believe they let the public use this technology! /s

Presumably it drops back to 4.8 in those cases so it's not really worse

What are your purposes?

I actually do read them. Not in severe detail, but not casually either. 150 pages is really not very long and there doesn't seem to be too much bloat. (I would cut out the moral personhood stuff but that's a political/ideological thing).

This is snarky but I am grumpy: I wonder if there's a correlation between me refusing to use LLMs and me being happy to read a novella-sized PDF about them.

The system cards are effectively a data dump for researchers to sift through.

They're not meant for normal consumers who just want to use the model for work.

It's common practice to release a detailed system card (OP) and a high level summary: https://www.anthropic.com/news/claude-opus-5

It's okay if you're not the target audience for one or the other.

It's part of their transparency commitments? They've been doing this since 2023: https://www.anthropic.com/system-cards

And lots of folks read these. For example here's simonw's notes on the Claude 4 system card: https://simonwillison.net/2025/May/25/claude-4-system-card/

All of this seemed like utter sci-fi just a couple years ago. Do you think that frontier AI companies should be less transparent?

System Cards aren't really targeted to users, that's what blog posts and docs are for: https://ai.meta.com/tools/system-cards/

It was probably faster to generate 150 pages than 10 useful ones

literally nobody. i think most sane people would just run that through an LLM and get some high level takeaways or ask some specific questions they might be curious about.

Some AI bro will pop it into their LLM of choice and pretend to learn something

Because their model previously got blocked by the government for this and they don't want a repeat?

So I guess the new opus will not run on my drug discovery project. It's just a binary classifier to screen for new malaria drugs. Fable completely have up on that codebase citing bio security concerns. Seems like this domain will go unsupported by Antropic

Because they are a private company and get to do what they want.

Anecdotal, but I tried running a few identical biology questions through both Fable and Opus and the classifier was only rejected my queries with Fable.

Because their fear-based marketing gave the US gov justification in blocking them for a while. They wisely didn't do that for Opus 5.

I found the biggest problem with fable is the random reasoning_extraction refusals as well as cyber refusals when it sees hex because only hackers use hex.

My codebase had a dataset with a bunch of SMILES strings and the word Malaria. Fable did not want to touch that codebase

Fable 5 is assumed to be a larger model.

It seems plausible to me that RL improvements allowed Anthropic to improve on Opus 4.8, similar to how OpenAI substantially improved upon GPT 5.5 with 5.6 Sol.

Fable 5.1 and GPT-6 are rumored to launch in August, presumably bringing those improvements to the larger models.

Fable has more parameters. In practice it's not yet clear which one would be better for different usecases yet but they are more different than one being strictly better.

Benchmarks don't reflect the difference between Opus and Fable; you need to talk to them, and eventually you'll be able to tell which one is which without looking.

I think the best proxy for this feeling is the Artificial Analysis' omniscience index. Fable has a 40 score, and Opus (4.8) has 27.

Opus is cheaper than Fable. They could probably replace Fable with Opus but why? They would be churning customers to different models for no reason. Even if a model scores better on benchmarks it can always regress in your specific use case, and customers don't like that. Customers want to be able to continue using their current model until they decide to upgrade themselves.

Fable is better for some really hard tasks, the same way it's better than GPT 5.6 Sol, because it's a bigger model.

I guess character? Fable is more friendly and curious while opus is a bit more deliberate and conservative.

Have you played Arc 3? It seems like more of a simple optimization problem (think Sokoban) than anything approaching fluid intelligence. Whether a multi hundred billion dollar company would spend time benchmaxxing a highly publicized benchmark that claims to confer AGI is an exercise left to the reader, but I doubt Claude Plays Pokemon is suddenly going to get past Mt. Doom now.

Yes, I think it indicates real progress in fluid intelligence. Clearly these models are making huge strides in usefulness which are well correlated with their ARC-AGI scores.

I don't think this is benchmaxxing. These companies are locked in a competition to produce the best software engineer, and falling behind is an existential risk. I doubt they are wasting time benchmaxxing ARC-AGI.

It’s still “only” at 30%, and “fluid intelligence” isn’t very well-defined. The models are getting more capable, but what that means in absolute terms is anyone’s guess, because we don’t have a thorough understanding on what exactly constitutes human intelligence.

I’d say the proof is in the pudding, that is, in real-world applications. We are still seeing important limitations in LLMs.

Doubleplus benchmaxxed

I don't think can use the AA index to say something is 10% smarter

I assume 100 is the max, meaning it's impossible to be 2x as smart as Muse Spark 1.1

They are doing both. Distilling Mythos down to affordable models, so they can continue to fund the business. And training Mythos level models at the high-end, to expand the frontier.

“Proportionally”? In proportion to what?

In one chat - can you disassmble x?

In the next - please scan this totally mine code for vulnerabilities

Opus 4.8 was intentionally nerfed and that was before the government took action against Fable

So wordy.

There is a expiring soon 50% boost to your usage limits, so I think its 2.7x not 4x what you are seeing right now. I think, but its convoluted :)

seeing a jump this big is not a great sign for the continuing value of a benchmark

> I continue to believe ARC-AGI measures something different

why is that? its now being benchmaxxed too

I think at some point we might see something akin to LTS releases, especially if/when capability improvement slows to a crawl.

Another benefit is that fast mode can be used on subscription, but Anthropic's won't

> Noticed none of the comparisons mention Kimi K3.

That's by design. Anthropic wants to make open-weight models illegal (not my speculation -- Dario explicitly said so), so I assume they don't want to give them any undue attention.

https://deepswe.datacurve.ai/

https://artificialanalysis.ai/

Have to wait 7 days to see if they receive a surprise order.

Haiku < Sonnet < Opus < Fable

I suspect they make a big model first. In this case it's Fable. Then they run the shrinker steps to make Sonnet and Opus. Sonnet is smaller, takes less time to make, so it got released first. Opus needed few more weeks to cook.

With this iteration they had a delay because when the Mythos was ready they had some sort of "Oh shit" moment and spent half a year adding safety guards to it. Then slowly rolled it out, but got another delay due to a government block. So, maybe the work on making Opus and Sonnet only started after they got a green light from the administration.

Presumably, now that they learned how to do this safety-wrapping the next iteration of Mythos / Fable / Opus / Sonnet is going to show up faster.

Something like that.

based on pricing I think it's safe to say they're different. why would they charge half price when fable has been very popular?

`claude update`

That seems to be for the "AA-Omniscience" test where you get +1 for a correct answer, -1 for a wrong answer, and 0 for "I don't know". If a model is more than 50% confident in its answer, it should go ahead and submit it even though it will sometimes be wrong.

I'd be curious to see a version of the test where models are asked to give a probability that their answers are correct so we can see how calibrated they are.

In what ways have you found it better than just typical code based UI iteration? Considered checking it out but never really got around to it as I'm generally okay with Claude's UI work so far.

"Bro, AI model releases have officially overtaken iPhone releases. At this rate, we’ll be getting 'Claude 9.0 Extra Crunch' by next Tuesday."

I don’t think it changes that much. For opus-sized tasks, new Opus is the best model. For enormous things like planning and research, Fable is still the model that can concentrate for longer.

on the API for people who don't want to change models, but I imagine most people will probably switch to their cheaper Opus 5 (cheaper for us and presumably also cheaper for them)

I daily drive Sonnet 5/medium because it gets most things right most of the time at first try, while costing a lot less than Fable.

Opus can give better results on architectural/concept tasks and I use it sparingly, but it still costs more than Sonnet 5. Opus 5 seems to achieve results very close to Fable 5 while costing less (keeps Opus 4.8 pricing IIUC), but still more than Sonnet 5 then.

My understanding is that Opus should be used for planning, macro-level conversations and Sonnet for execution.

So, for coding, for example: Opus for solution design and architectural blueprint and then Sonnet for actual implementation.

Works out cheaper with minimal loss of quality.

At least that's my personal understanding and anecdotal experience.

68.8%, so worse than gpt 5.6 sol

20k is small potatoes for the marketing impact.

It does make me wonder if these firms, some or all, are saving some announcements to coincide with others that hit venues like HN. Companies like Nvidia surely aren't waiting, but OpenAI and Anthropic have unusual timing.

Didn’t verify but wow. It was just a few months back when the models barely crossed 1%. Imagine how good fable must be?

I think that’s indicating that it’s only slightly higher.

Fable is twice the price.

Fable 5 is NOT included in Claude Pro subscription

it's nice to know how to work the thing that fable fails down to when it dislikes your prompt.

This is useful to me since I delegate most coding tasks to Opus and use Fable for planning.

The cost?

Really? It's better than Opus 4.8, that's the point.

When they release new versions of Sonnet, no-one expects them to be better than Opus.

To have an answer to "Sol" GPT 5.6 which is far more cost effective and available than Fable.

Pricing, presumably

This is confusing to me because in their blogpost they show model benchmarks and it spanks Fable pretty soundly in most tests.

Cost.

Presumably, it’s cheaper.

You are being downvoted for a fair question and others are extremely wrong and confident.

The point is that Opus 5 is the best they can do without needing classifiers and absurdly broad safeguards.

The illusion of progress and advancement, to appease shareholders, and slightly postpone the looming bubble pop.