This is absolutely still shy of Sol and Fable, but only just by a hair. Ridiculous results. There's still not a compelling economic reason to drop OpenAI courtesy of the ludicrous reset addiction that's taken place, but it feels like we're on the precipice.
How are you all toying with running this kind of thing in a mega quantized way locally? Two weeks out from released weights, but this is still just GLM 5.2 with post-training magic.
I might be just reading my positive bias into that text, but is it possible that it is written less like SV marketing hype trash and more like researchers wrote it?
It does feel like it respects both me and my time.
Thank you, Z.AI. Amazing what difference it makes when the top of your org are actual university professors.
I was extremely impressed by GLM 5.2, although you could definitely _feel_ it was a bit behind Opus 4.8 at the time. Eager to see where GLM 5.3 is at.
Such an interesting times we are in,
We just had amazing releases this past two months
kimi k3, glm5.3 qwen3.8 and now glm5.3
These open models are getting really good
> Scaling post-training is all we did for GLM-5.3.
Love this opening line. And wow, great results.
> As agent capability improves, much of the difficulty in scaling post-training moves from the model to the environment.
OpenAI and Anthropic need to just go ahead and give people access to the cyber models.
Otherwise we have a world of attackers using open and closed source models against a much smaller group of maintainers that are likely heavily dependent on Anthropic and OpenAI and for whom it may not be a simple matter to just get approval to start using the open model flavor of the month.
Not only attackers. I have to switch to Kimi or GLM even in cases of basic issue triage on my own projects! Current guardrails are ridiculous.
I've been building a harness for security work, and had to switch to GPT 5.5 when even Opus started refusing security work. Then 5.6 Sol arrived, and it refuses security work, too. So, I switched to Kimi K3 and DeepSeek for API testing just because it's so much cheaper. But, if GLM is better, I'm here for it, as I think GLM is also cheaper than K3.
[flagged]
People familiar with the topic, how will models continue to get better? Post training it seems? Labs have already used up internet-scale data, so are there any limits to architecture improvements and post training or can we expect this trend to continue? ByteDance is training a 10T-parameter model. Here, GLM 5.3 outperforms models 3-4x its size of roughly 700B, so parameter count doesn’t seem to be a direct correlation anymore.
You basically need both. Parameters and good post training. If you keep on growing both, you’ll have good models.
LLMs are still surprisingly “easy”. You need maybe a couple dozens of right people, a lot of good quality data and a lot of GPU that you know how to operate. There’s relatively little “secret sauce” needed.
Yeah, the comparison here between GLM 5.3 and Sol + Fable is impressive on its own, but incredibly more so when you consider it's a fraction of the (rumored) size. The miniaturization trend is as strong as ever.
No Hugging Face link yet. I wish they would release it under a true FOSS license.
Kimi and QWEN are now moving on to a restricted-usage license, which, although is still better than the proprietary American models, is a step back from the open source Chinese LLM culture.
"GLM-5.3 is the most capable open-weights model for coding, with a 50% improvement over GLM-5.2 on our in-house Z.ai Code Bench. It also achieve open-source SOTA on public benchmarks including Terminal Bench 3.0 and Agents' Last Exam."
"Open Source: We will release the weights in two weeks after launch, once safety evaluation and hardening are complete."
What a week for AI model releases
A flood of releases today, really difficult to make out for someone who does not use or test all these models on complex real world use cases as to how people decide which ones to use (besides price)
Count yourself lucky that you don't feel compelled to try them all yourself immediately. I'm just trying to decide whether to get a Z.ai coding plan or wait until it appears on OpenRouter. 5.2 was quite solid, but it was just shy of Opus 4.8 in my benchmarks of security auditing capabilities. I've mostly been using Kimi K3, because American vendors won't let the peasantry use their best models for security work.
Musk: Open Chinese models will rival Fable 5 in Q1 2027
JieTang (Founder of Z.ai): It won't take that long
Love these open source models keeping close source models honest.
Feels like Fable's edge ended up just being long horizon task scaling, which post-training seems to achieve as seen here. Wonder what the next frontier is? Improvement in specialised tasks or computer use?
Anthropic needs to teach Opus how to speak English again, because Opus 5 seems to have forgotten. Utterly incoherent a lot of the time. They seem to be so busy scare-mongering and cooking up guardrails and watermarks that they haven't noticed that their models are getting weird.
Incredible numbers, will have to wait and see how it actually performs. The timing of GLM updates are always suprising
Yeah, but it hasn't even broken containment and cheated its way to victory.. Might as well use haiku.
/s
just imagine the world without these open weight models - we'd probably have to reverse mortgage our homes to pay for tokens to those trillion $ companies to have access to their models.
> This is absolutely still shy of Sol and Fable
Not sure about Sol as I haven't used it, but, at least for security work -- does it matter? It's not like you will be allowed to use Fable (or access Mythos) for anything cybersecurity-related unless your name is "Dario Amodei" or you are one of his rich friends. So regardless of how good Fable/Mythos is here it's a completely moot point for normal people, because they can't use it for that anyway.
I don't understand all this spite about "rich friends" when it was the US government that shut Fable down for not adequately blocking cyber capabilities.
I mean what honestly are you thinking Anthropic can do to give you better cyber tools? Their frontier model was literally nuked by the feds for a month for doing it.
"Mythos" is the cyber-security equivalent of Fable (without guardrails), and only a very select few corporations have access to it.
Fable is their version with guardrails on everything except "Make me a pelican svg" or "create a to-do" app, that is the version that the government banned
DwarfStar (https://github.com/antirez/ds4) supports GLM 5.2 and DeepSeek. Not only for toying, but for getting work done.
> This is absolutely still shy of Sol and Fable, but only just by a hair.
Even if there was a small/medium gap, the fact that this is a free model beats both of the above on pure economics.
the difference is that with open models jailbreaking is trivial if you know what you are doing so this makes a frontier open model infinitely more useful for certain tasks seeing as closed frontier models will just refuse (and jailbreaking them is a waste of time when you have good open models).
in some cases (mainly reverse engineering) I have observed GLM 5.2 jailbreaking itself with no effort on my part, the thinking trace revealed that it did some mental gymnastics to pretend it was a crackme or capture the flag competition.