Join the discussion
Write your take first — we'll ask for email only when you're ready to publish.
- Hacker News
- Whatever it is, it certainly crosses the threshold that lets it follow instructions adequately and do real work, so I'm just letting it work on my personal projects and save some money
- I hope it is glm air. We need more "small" models. Big models are more capable and useful, but for majority of tasks some smaller models can work just fine.
It is funny that google gave up on this market, leaving the whole price range to Chinese models.
by npn - Chinese labs are not even in competition mode yet. You'd be seeing them paying you for each token you use when they are in that mode.by tw1984
- Sounds like this could be GLM 5.3 vision. Reports said that outputs were identical to base GLM 5.3.by LorenDB
- Yeah, it's very likely GLM.
Trick: it's easy to check this - just try eg. a specific short almost-nonsense high-entropy phrase such as "Scarf Color Plump 蘼撅" via OpenRouter using the playground for different providers/models, and observe the shape and language of the reasoning and the response, which tend to be quite different (MiMo and GLM are quite similar, but still identifiably different enough).
by jaen - It's Chinese. Won't answer anything about Tiananmen Square but will gleefully give you instructions to perform various electronic warfare attacks that opus and fable instantly refuse.
Side tangent, why is fable so weird about questions involving "Welch's method"? Even really trivial ones it'll shut down frequently. CFAR and STFT are both totally fine but Welch's is apparently taboo, it's wild.
by fedpost - Does the Venn diagram of people eager to study history and the people stupid enough to use an unreliable chatbot to study history really have that much overlap?by t-3
- Rumors from other sources based on how it behaves it's mimo v3by walrus01
- I had the opposite experience. It happily discusses Tiananmen Square but said it would refuse to help with anything "malicious" like writing malware or phishing content.
- I suspect Fable refuses to talk about anything biology whatsoever (so much so that I can't use it) -- and Welch's test, i.e. an unequal variance t test, is beloved by biologists. Stupid, but there ya go...by azalemeth
- Model is suspiciously fast and has a low reported output token count (using via OpenRouter's Chat), both of which aren't representative of models from the big Chinese labs. Odd.by minimaxir
- > suspiciously fast
They're reporting ~30tps, that's about in line with many medium sized models served by Chinese providers
by drbscl - GLM-5.3 is one of the faster models, at least according to artificialanalysis - openAI and Anthropic are the slowest.by nkmnz
- > Model is suspiciously fast
> aren't representative of models from the big Chinese labs
There were reports that China has let Nvidia's chips through, so this might be it. Testing both the chip and infrastructure.
by re-thc - I highly recommend feeding all your proprietary data and confidential personal information into this model as quickly as possible. What could possibly go wrong?!
In terms of equivalence of suspicion, this is the external inference provider equivalent of getting free steak that was smuggled out of a grocery store inside somebody's pants.
by walrus01 - Just a few years ago, people would lose their marbles if some software installed a background agent to send you notifications or something.
Now we agreed that it's totally normal to have software that does remote code execution on our machines, for which it first has to transfer all data to a remote server.
We've already normalized this. It doesn't really matter who gets access to said machine/data - they will all retain information, and they will all train on it, regardless of what user agreement sais. It's not like OpenAI and Anthropic didn't train on things they didn't have permission to train on.
by glub - You say it like it's less dumb to feed this kind of data to other EU or US models.
- We’re about a year and a half past this conversation. The industry has settled on “Don’t use it for work, unless your company is okay with whatever models. Everything else is whatever, super-majority really does not care at this point.”.by tokioyoyo
- All of my non-work AI coding is that open-source, so I'm happy to feed my data into the machine.
It's a win for me: my code goes into the training data, and my sessions are fed into future training data, making the model stronger at the type of work I do.
- > I highly recommend feeding all your proprietary data and confidential personal information into this model as quickly as possible. What could possibly go wrong?!
What is special about this model? The model's provider is not anonymous. OpenRouter knows who it is (and apparently decided that, in whatever way they always do it, it is okay to work with them). Using this seems roughly equivalent to using any model through OpenRouter, as far as I can tell.
Or is this just meta-critique?
by jstummbillig - I'm kind of fascinated by how many of the same audiences who are highly skeptical of OpenAI and Anthropic are the same people running straight to other country's models.
The most oft-repeated rebuttal I've heard is that they don't care what other government know about them. I guess their threat model hasn't considered any privacy issues, data mining, or leakage risks, just the possibility of the federal government doing something to them?
by Aurornis - There are low stakes use cases where this kind of stuff just doesn’t matter. Not every use case for an LLM involves sensitive or even non public data.
Eg. I have a need to search transcripts of published recordings to extract entities for tagging purposes, find semantic shifts for chapters and other things. The underlying content is already published. If they want to train on my prompts, that was something they could have done with no issue and minimal effort anyway.
Sometimes you don’t need to care why the steak is free.
by dghlsakjg - I don’t understand this mentality which I see over and over again on here. It’s not like the US frontier labs are beacons of morality and transparency.
It’s also pretty accepted within this community that a lot of data fed to US tech companies ends up with the Israeli government.
Meanwhile the current US government headed by Donny Tango has done a very thorough job of proving itself to be about as predictable and dependable as a rabid dog on crack.
Throughout the events which have transpired since a certain orange charlatan took office it is objectively true that the Chinese government has portrayed itself as a much more stable and sane entity.
We really need to stop this elitism and recognise the reality.
by makingstuffs - On softer/looser/creative matters, this is an extremely impressive model. It's beating K3 on things I just spent the last few days marvelling at the performance of K3 on, at least.
Visual reasoning is not great (unsurprising).
by gadtfly - Based on its indecisive and far-too-lengthy thinking traces when given complex instructions that span system and user messages, as well as a rudimentary stylometry (POS ratios in thinking traces, mainly) comparison with latest non-stealth models, this is almost certainly a GLM model.by spdustin
- Wasn't the last "big" stealth model glm5.1?by walrus01
- The "mia" persona on X has a specific vaguepost:
https://x.com/MiaAI_lab/status/2090736338328748220?s=20
> "I've got a confirmation on what model is Ox Alpha, but I can't share it yet. What I can say is this: You should ALL get really excited for this one!!! And it’s NOT what you think it is"
And others have said they have done analysis and found it to be GLM 5.x related.
That said, Mia said "it will be OSS" and "it'll run on 2x DGX Sparks" - well GLM 5.2 can run on 2x Sparks, but slowly and heavily degraded (quant). So doesn't really confirm/deny that suspicion.
by alexellisuk - I'm suspecting it is Deepseek flash-vision-exp V4by estebarb
- It did an absolutely terrible job at generating CSS, where I instructed it to finish implementing a bright and dark theme based on a palette through the use of `color-mix()` and it just went ahead, removed everything I pre-added and replaced it with hardcoded hexadecimal color values.by hxii
- Yeah, whatever it is, it's particularly bad at front-end from what I have been seeing.by gexla