upvote
It is only a price drop if price * tokens used is less
reply
Happened to be testing a "review patches on a mailing list" harness I was developing; here are a sample of the latest results, testing 12 patches containing a total of 14 issues:

Opus 5.5: Found 8/14 issues. Total cost: $15.40

Fable 5.1: Found 7/14 issues. Total cost: $66.34

Opus 5: Found 6/14 issues. Total cost: $15.19

Sonnet 5: Found 2/14 issues. Total cost: $19.15

This is a relatively small sample size, but it was both the best and the cheapest.

ETA: NB this is "Equivalent API" cost as reported by claude's CLI; I was using my subscription.

reply
They're claiming a drop in token use too, and that it nets to 40% cheaper.
reply
Unfortunately, they're full of it https://artificialanalysis.ai/models/claude-opus-5-5#token-u...

It does work out to be a similar cost per task though

reply
You should probably look at the cost/score graph by effort level instead:

https://artificialanalysis.ai/models/claude-opus-5-5#intelli...

It is most of the pareto frontier.

reply
Not disputing the increase in quality, just stating that non-cherry-picked benchmarks show it is more verbose at Max effort
reply
so don't use it at max? The benchmarks suggest that high/xhigh are more than sufficient to be ahead and a whole magnitude below max with regards to token usage. I'd treat that as an outlier and not how verbose the model is in general (QED I know)
reply
Is verboseness the only measure of token efficiency towards overall task completion?
reply
Disagree. Our internal company tests showed a cost per task drop from 0.35usd to 0.16usd . Opus 5low vs opus 5.5 low
reply
Very fast you were.
reply
Even created an account to tell us just that.
reply
I don't think so, I typically use Opus 5 on High, and 5.5 scores lower on token use:

https://artificialanalysis.ai/models/claude-opus-5-5?models=...

reply
I tested it with Claude Code, and I can confirm it's way cheaper, better, faster and less verbose than Opus 5.
reply
parent means that they could get more client / a larger part of the market, which would lead to more income (more tokens) despite lower marginal prices
reply
Claude adapts to OpenAI’s surprising move to simply deliver better performance than Fable 5.1, better tools as well as featuring very low pricing.

Fable 5.1 literally was a money grabber. While I liked the results, tokens were burned so hard it was embarrassing, while Astra seemed to not care.

Also Claude makes it very hard to pay for additional token budgets, allowing only credit cards. I don’t use mine anymore since I don’t need it in everyday life I was dumbfounded.

So Anthropic is just copying OpenAI so to say, matching them and essentially with Opus 5.5 being Fable 5.1 in disguise, all they do is reduce costs.

Competition works.

reply
People are paying for Opus 5? Not just burning down tokens left after they enjoyed Fable on the sub? Amazing.
reply
My workplace doesn't even offer Fable. And on the sub, I've had a hard time understanding Opus 5, but Fable can deal with it with subagents.

If 5.5 is any better, I might try to do agentic-assisted development instead of just telling fable to delegate

reply
Telling Fable to delegate is agentic development. At least I thought so until reading your comment.
reply
When you need to disable data retention, you cannot use subscription plans.
reply
We can’t use fable at work, opus and Astra are as good as it gets.
reply
Enterprise and most Team accounts use API pricing, they don't have an included-usage quota.
reply
Enterprise, and APIs
reply
We haven’t been able to use opus as much as we’d want because it’s been too expensive for general use, price drop is good so I can stop juggling different models and just use this daily unless it has some weird new issues
reply
Speaking for myself, I have not been able to use Opus as much as I’d want because its verbose prose makes human reviews of its assumptions, architecture proposals etc. more painful than its predecessors. If they’ve solved that, I’ll be accelerating through my backlog that much faster, and using tokens accordingly.
reply
Price per task (not per token) is what really matters.
reply
Agree. But similar to how ISP use 200 mbps (bits) instead of 25 MBPS(bytes), i think this trend isn't going away.
reply
That analogy doesn't hold; at least w bits vs bytes it's still "data over time".

In this case it's measuring something nearly meaningless. You could charge 100 times less per token, but if task completion takes 1,000 times as many tokens, it's not much of a bargain.

reply
>and potentially about Anthropic future profitability too

have they ever shared anything about their revenue mix between consumer plans vs per-token billing? this is a revenue cut on their API billing, but they're not saying anything about increased limits on the plans. so all the plan revenue just got more profitable.

reply
All the anti-AI people constantly say that any moment now the prices will skyrocket and in the end human work will be cheaper compared to using AI.

It doesn't look like that's happening, on the contrary the prices are falling especially when taking into account capabilities.

reply
It's the Chinese open source models. They're barely behind the frontier, making AI a commodity, forcing openAI and Anthropic's margins downward.

I'm hardly a fan of China/Xi, but I do appreciate and benefit from this.

reply
China doesn't care about money. Imagine a world where it's globally normalized to ask a Chinese LLM who to vote for, what happened in Hongkong, about the Uigurs, or if Taiwan is a country.

They will burn as much money as necessary to make that happen. And they have a virtually infinite amount of liquidity.

reply
FYI the United States do not recognize Taiwan as a country.

There's only 12 countries that do on the planet, the most "relevant" of them being Guatemala and Haiti.

As or the LLM topic: you can download weights of chinese models and remove any censorship and bias. Can you do so with american closed ones?

reply
Footnote on their pricing page says:

> Cache hits and refreshes on Claude Opus 5.5 are priced at 0.05x the base input price.

If they do the same for Haiku and Sonnet 5.5 then we should also see 5c/mtok and 10c/mtok cache read for those models, respectively. Still too high for Haiku IMO, Luna is 2c/mtok.

reply
Since something like 98% of tokens are cache hits, that's a pretty substantial drop from 0.1x
reply
6-Luna dropped it to 1c since you wrote this! :D
reply
60% cache read is cool, but subscription only get 25% more according to Cat. I'm confused
reply
I think gpt 5.6 family also dropped pricing but didn't give any more usage for the subscriptions. Maybe it's a way to silently lower the value given to subscriptions while keeping API pricing competitive
reply
25% more usage sounds about right given the other token costs are down about 20%? I don't think cache read is a big portion of the overall cost.
reply
Anything with long context quickly gets dominated by cache reads. Especially for interactive sessions I’ve got cache read % between 95% and 98%.
reply
In my mix it's usually 98% or 99% at which point Fable 5.1 was pretty close to the same cost as Opus 5 due to the cheaper cached read. I've seen similar numbers for other people with long-running tasks running experiment loops and than sort of thing.
reply
> I don't think cache read is a big portion of the overall cost.

For long running tasks it is. That's what made Deepseek so cheap.

reply
Yeah, flash models, DeepSeek, MiMo, GLM, I love those things. For simple tasks like a daily routine shit, just setting up stuff and then doing the hard stuff in Claude/Codex, that's a reasonable approach for someone like me, a "gentleman code farmer", lol. And even lower tier stuff, I have the local models taking care of. Now that Jev is out I can finally have a true AI sysadmins managing my "cloud in the basement" homelab at the cost of electricity, which is not cheap btw
reply
source link please?
reply
This is good. Probably to match GPT pricing, though frontier Claude models are still not as token efficient.
reply
"Opus 5.5 requires less compute to serve than Opus 5, and its pricing reflects that."
reply