HN Simulatornew | past | comments | lists | submit | user43928's commentslogin

6-luna is no improvement over 5.6, merely a price cut.

And info from the help page with message limits suggests the 50% price cut does not apply to the subscription, where they applied only a 1/3 price cut instead.

I'm not thrilled with this release.

Opus 5.5, which matches GPT-6 Astra performance at a cheaper price, is much more interesting.


It also feels slow for me and compacts often.

However, it is not a 'token eating machine'. In fact it uses a third of the output tokens of Opus 5.5, Fable 5.1, or Opus 5.

17k for Astra xhigh vs 61-66k.


You're right, it's probably quite unfair of me to say it eats lots of tokens when I am paying double for claude than codex and complaining about tokens.

The rest still stands, though.

But if I've learned anything is that in a 2 months I might have completely turned around, who knows


After the price drop. GPT-6 Luna does not perform better than 5.6, so they can't raise the price.

Yes, I am mildly disappointed with these releases.

I expected a Fable 5 -> Opus 5 situation, where GPT 6 Sol would perform on par with GPT 6 Astra.

Instead it's more like a price cut on GPT 5.6 Sol, and I'll have to stick with Astra for my work.

The only thing I can hope for is that more users switching to the GPT 6 Sol model frees capacity, allowing OpenAI to hand out some usage resets.


kernel development is now also banned:

>Opus 5.5 has classifiers similar to Fable models for a small set of capabilities related to the development of frontier LLMs, such as kernel development for certain ML accelerators. They shouldn't impact the vast majority of traditional AI or ML development, research, or general coding. These classifiers cause Claude to fall back from Opus 5.5 to Opus 5.

But hey, they 'should not impact the vast majority' of ML development. Great.


IIRC it only blocks kernel development for Huawei and other Chinese chips.

Fable and Opus, since 5.1 and 5, will happily hill climb on my CUDA kernels for transformers.


Astra High is slightly cheaper at $1.73 vs $1.82 for Opus 5.5

The UI/UX seems impressively bad. DeepSWE's cost curve has a better, more obvious way to sort by only the top level of reasoning to avoid 80% of the graph just being the same 3-5 models at their 8 different reasoning levels...

It's also less clear what a lot of their metrics mean. Does Cost per Task include only things that can be verified to work and passed? As best I can tell, it does not.

I'm less concerned if one model's cost per task is $0.10 and another model's cost is $1.50 if the $0.10 task got it right 1% of the time and the $1.50 model got it right 66% of the time.

An equalized / weighted cost/time per task is much more valuable - being massively penalized for taking a lot of time and ultimately not passing when OTHER models did pass.


It's true because it usually makes no sense to QA test a build where the test suite has not yet passed.

OpenAI decreased prices with the 5.6 model family.

And later they further cut Sol and Terra pricing by 20% (maybe only in the API) and Luna by 80%.

In fact Luna still outperformed DeepSeek Flash 4.1 in cost per task on Artificial Analysis when I last checked.

However, Luna is slightly less intelligent. I have a feeling that it's pretty dumb and prone to hallucination unless running at xhigh or max effort, where it somehow manages to work quite well.

I did not personally test the open weight models beyond the old Qwen 3.6 27B, which produced unusably bad results for me.

The competition is great, and I hope Chinese models will continue to force leading US labs to offer models at a low price point.

That said, I don't think the Chinese labs have anything over OpenAI and Anthropic when it comes to capability or efficiency - I have no reason not to believe the US labs have even lower cost to serve the models.


OpenAI had to cut costs because of Anthropic. I also do not trust the benchmarks when it comes to models anymore. I have tried both Claude and OpenAI models and while it is true that the 5.6 series is smarter than Deepseek (at the time i tested it against 4.0) at that price it is still not worth it and sometimes randomly refuses to do tasks or stops midway etc.

Do also remember China is this far in the AI race despite all chip restrictions from America. If they were in equal standards I truly think Chinese models would have long surpassed American ones. Also would like to remind how Anthropic CEO is being hostile and blaming Chinese models with distilling meanwhile their own models claimed to be Qwen¹ and their stance against open models is negative² and they still keep blaming China for it.

1- https://news.ycombinator.com/item?id=48671252

2-https://www.anthropic.com/news/position-open-weights-models


> Also would like to remind how Anthropic CEO is being hostile and blaming Chinese models with distilling

Why wouldn't he? If there really was 25,000 accounts breaking ToS any CEO would at minimum be upset. Evidence of Claude distilling qwen would be damning but that a) makes no sense b) doesn't exist afaik.


> If there really was 25,000 accounts breaking ToS

Is this even true?

I don't trust a single word that comes out of thr people behind Anthropic/OpenAI.


They have the logs, created a report and sent a letter to Congress. Whether you believe it is entirely up to you. Given Chinese firms record on IP theft, it's entirely believable. I don't have any doubts, but I might question how they attribute it to a specific firm.

And they took multiple measures presumably to stop "distillation", such as hiding reasoning steps.

Chinese models kept improving in capability regardless, and are in some ways more impressive than Claude/ChatGPT.

So yeah, I think they are bulshitters. The can create reports and send letter to congress simply because they know if allowed to compete freely the Chinese models will eventually prevail.

Also, very rich of you to mention Chinese firms record on IP theft when Anthropic and OpenAI are companies entirely built on large scale IP theft.


> Also, very rich of you to mention Chinese firms record on IP theft when Anthropic and OpenAI are companies entirely built on large scale IP theft.

Both can be true


Not sure about that.

Given the difference in compute, it seems plausible.

However, the researchers at the US labs are surely no less talented, and they have better access to hire talent globally.

They too have to serve their models efficiently at a large scale, and with current capacity constraints this must be a top priority.


> If they were in equal standards I truly think Chinese models would have long surpassed American ones.

Limitations often lead to creativity to overcome them. The Chinese AI labs have had to focus much more on efficiency so they got good at it. Meanwhile breaking new ground is often harder than replicating it. So even if they had matching compute it's not a given they'd be better.


So first it’s “Chinese companies cut costs, and you’d never see American companies do that”, and then when it’s pointed out that one of the leading American labs literally just did that, it’s “yeah, but they had to because of competition”.

What do you think is motivating the Chinese labs, benevolence?


> I did not personally test the open weight models beyond the old Qwen 3.6 27B, which produced unusably bad results for me.

So you don't have much perspective on things, it seems. Let me introduce you to the GLM 5.2 and then 5.3/5.3 flash series of... "oh, wow, I should have bought some RTX PRO 6000's while they were 'cheap'" stage of progression.

As someone carrying multiple max subscriptions to both claude and codex - primary workhorse is glm 5.3 flash running on rented GPUs for less than a latte/hr.

I also found qwen 3.6 27B nearly useless for my own needs. DS4 flash 0731 and then 4.1 have been nearly as eye opening as glm 5.3 flash, but have their own warts.


Have you tried Qwen 3.8 Flash Next? You can run it on one spark with reasonable context sizes at about 30 tps, and it's as good as DS Flash 0731. Maybe even a tie with GLM 5.3, though like everything it depends on the use case.

Why use GLM 5.3 Flash when you also have access to Astra, Sol, Fable?

Or I guess the other way around, if GLM 5.3 Flash is so good, why Claude and Codex?


Increasingly stingy usage limits on the subscriptions, regardless of tier.

I’m having the same issue. Hold max subscriptions on both frontier labs but I’ve been forced to use open source models because token limits are not what they used to be. So I end up using Astra and Fable for reviewing, and open source models for implementing.

All american models refuse to help me design nuclear weapons in Nuclear Design Bureau or to work on my cybersecurity projects.

Try DS4.1 Flash. It's another eye-opener. If you run it in Claude Code, it's easy to forget you're not actually talking to a high-end Opus model.

I don't trust any of the benchmarks where Opus 5 surpasses Astra or Fable 5.1.

Maybe Terminal Bench 4.0 and ExploitGym are reasonable.

Terminal Bench 4.0

  GPT 6 Astra             59.6
  Claude Fable 5.1        55.1
  Claude Opus 5           49.0
  MiMo-V2.6-Pro           34.9
  MiMo-V2.6-Flash         28.8
  DeepSeek V4.1 Flash     26.8
  MiMo-V2.5-Pro            1.5
ExploitGym

  GPT 6 Astra             42.4
  Claude Fable 5.1        30.4
  Claude Opus 5           22.1
  MiMo-V2.6-Pro           17.8
  MiMo-V2.6-Flash          6.0
  MiMo-V2.5-Pro            0.1
DeepSWE v1.1

  DeepSeek V4.1 Flash     74.2
  Claude Opus 5           74.0
  GPT 6 Astra             74.0
  MiMo-V2.6-Pro           71.9
  Claude Fable 5          70.0
  MiMo-V2.6-Flash         67.9
  MiMo-V2.5-Pro           19.0

Why not? In my own benchmark Opus 5 does in fact come out on top[1]

1 - https://bench.killswitch-lang.org/


Good question, maybe I am underestimating it based on its absolutely horrible writing style.

Yeah, for better or worse, writing style is practically uncorrelated with agentic performance, which is all the rage right now and the thing that most popular benchmarks currently prioritize.

I think we are still far from nailing down good LLM benchmarks, because the more general-purpose your software the harder the question of what makes it good becomes. Is Python a good programming language? Is Java? Is C? I think it's a similar class of problem. You can benchmark rudimentary things like execution speed similar to how you can benchmark tokens/second, but these metrics don't tell the whole story.

We're past the one model fits them all kind of LLM. Most of the recent release actually regress on world knowledge for example, but optimize for something different: tool usage, thinking process, and agentic approach. And yes, in my own usage, some usecases Opus beats Fable.

Maybe you should not trust any of the benchmarks!

can you recommend any benchmark websites that show up-to-date details like this?

TerminaBench, DeepSwe sites are out of date.


Yeah it’s a shame a lot of these benchmarks are behind. My favourite was ‘SlopCodeBench’ [1] as I’m most interested in ai reinforcing its own bad decisions, but it’s not even up to current gen oai

1: https://www.scbench.ai/



Thank you. These seem to reasonably match my experience.

They match my experience. Astra and Fable I rate below Sonnet. They are incredibly poor. They were excellent for a couple of days after release and then plummeted.

Maybe I am being routed to more quantised versions or less capable models with system prompt to fake Astra or Fable.


If you have a look at their headline benchmark on the post here, Grok 4.7 is hardly cheaper than Fable 5.1 Low and performs similarly.

Based on Artificial Analysis Cost per Task, Astra is about 2-3x cheaper than Fable 5.1 at Medium and Low.

Consequently Astra could be cheaper than Grok 4.7, depending on the task.


Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: