250 comments

[ 1.1 ms ] story [ 83.2 ms ] thread
Very interesting to see such a focus on cost for performance here
What really impress me is opus 5 is better in alignment than fable 5!
> Claude Opus 5 is not more capable overall than our most capable general-access model, Claude Fable 5

Ok then so what's the point?

Very impressive headline benchmark numbers. I expected a step change, but not past Fable. That said - it all depends on whether the classifiers make the model unusable...
Interesting timing to release this on the same day Jensen makes a statement on open source AI.
so almost fable 5 with 50% cheaper cost? sign me up
Rather interesting that this makes sonnet 5 look even worse! There is no reason to use sonnet over opus with low or no reasoning at all.
Great that there's a new model but they could fix their existing infra. We're considering dropping our Claude Team sub cause it's unusable recently. Constant bugs, dropped sessions, issues switching models, http errors. It's becoming ridiculous
Excited to use it? Will we be seeing Haiku 5 next? /s
I unironically hope Haiku gets an update considering it came out in October of last year and it seems like Anthropic just kind of forgot about it.
Models benchmarks start to get saturated again!
> Opus 5’s safeguards match those of Claude Fable 5’s, with one change: it now permits source-code vulnerability discovery at all access levels. This means that the model can support defensive cybersecurity work while still blocking vulnerability discovery in compiled binaries, which is more commonly used offensively

Why can't they also allow Fable to do so also? Why is source-code vulnerability discovery limited to a lower capability model? If Fable and Opus have the same safeguards, except for this one change, I see no reason they can't also allow this for Fable.

Noticed none of the comparisons mention Kimi K3. Is there a comparison chart?
Same cost as 4.8 but better that 4.8. Happy to get more efficient model. But is there any reason all companies are releasing models back to back after GLM 5.2.
> Opus 5’s safeguards match those of Claude Fable 5’s, with one change: it now permits source-code vulnerability discovery at all access levels. This means that the model can support defensive cybersecurity work while still blocking vulnerability discovery in compiled binaries, which is more commonly used offensively.

Okay so it’s worse than Opus 4.8 for my purposes I guess?

I don't think there's been any official confirmation, but even Fable safeguards seem to have gotten quite a bit of tuning, is less trigger-happy and less like regex matching.
I am very confused about what the difference between Opus 5 and Fable 5 is now. What is the purpose of having two models that are so similar? The main differences I see are cost and marginal capability, according to the Anthropic-provided benchmarks.
Looking at all these releases it’s not a surprise that model routing is the fastest growing segment in AI right now.

There are 10+ LLM companies, each with dozens of models of different modalities, each model with multiple size variants, then different “thinking” levels, then agentic modes, “pro” modes, a “fast” option, standard vs flex vs batch execution. And of course each end combination has a different input/output/cache token price.

Companies that say “give me a prompt and I’ll route it to the most ideal and cost effective model and setting for you” are capturing a ton of value from a gap that model developers don’t seem to understand exists.

lol I had to get ChatGpt to explain to me the difference between 5.6 sol, 5.6 Terra, 5.6 Luna, 5.5, 5.4 mini, 5.3 spark, and then there is low, medium, high, extra high, max, ultra, and pro… I still don’t really know, it feels like ordering hot wings.
I am not convinced this will end up being a domain of the ‘routers’ vs the clients, as in harnesses themselves. Thoughts?
How does model routing work if the prompt is static? Do the routers attempt to tweak the prompt to make it the best version for that model? For example, reading the “migration” guides for Claude Opus 4.6 -> Opus 5 is exhausting. Adding in all of the other models complicates it even more. Especially across vendors.
How does it perform on HuggingFaceExploit bench? Suspiciously absent, so not sure if I can take the model seriously.

On a serious note, I hope they improved their extremely sabotaging and unspecific bio safeguards, which prevented Fable from being used in any codebase that ever so slightly grazed medical terminology or data and made me switch to 5.6 Sol.

What is HuggingFaceExploit bench?
> . Opus 4.8 served as fallback on safety-classifier refusals for Opus 5 and Fable 5.

ffs just keep it man.

"Cybersecurity. Opus 5’s cyber classifiers are proportionally less restrictive than those on Fable 5. They allow Opus 5 to find vulnerabilities in source code, but block “binary-based” vulnerability scanning (a method more likely to be associated with malicious actors), penetration testing, and exploit generation."

Nice of them to be more explicit for what is blocked. Will be interesting to see if this is true or not.

Also, a notable lack of mention of open source models. They only compare themselves to ChatGPT.

From the prompting guide<https://platform.claude.com/docs/en/build-with-claude/prompt...>:

> Claude Opus 5's default user-facing responses run longer than prior Opus models'.

The benchmarks do show Opus 5 as slightly more expensive than 4.8, although the scores are much higher.

This still feels like a step in the wrong direction, though, especially with OpenAI making so much progress with the efficiency of their models. Fable's token efficiency made it seem like Anthropic would start following OpenAI's approach but that doesn't seem to have carried over to their other models.

Almost completely disagree. Slightly more expensive, but significant better on a per-prompt basis? For non-trivial projects, the former is a small linear increase, the latter is a (somewhat-)exponential(-ish) cost/time/sanity savings.
In my tests, it averages to much cheaper than Opus 4.8 on real tasks on account of being smarter and more token efficient.

I have a benchmark to build a game engine from a set of written instructions. It's a little tricky. Opus 4.8 did it in 470k tokens at a cost of $1.29 vs Opus 5 in 179k tokens for $0.33. (Fable 5 did it in 245k for $0.95)

Though if you really want to cut costs, Tencent's Hy3 model also got it right and did it in 283k tokens for $0.03

no infonat all about the default and recommended effort?