Comment by mordae
16 hours ago
No, they are not. They are winning because West is forbidden to use Chinese models for anything work-related.
16 hours ago
No, they are not. They are winning because West is forbidden to use Chinese models for anything work-related.
True, many people don't know GLM 5.1 and Kimi 2.6, really on par with frontier models. There's also Minimax 2.7, DeepSeek 4, Qwen, Xiaomi 2.5 Pro, etc.
China is leading in open source frontier models, so I don't really see how the US wins this one. At some point, companies and people will start running their own models in the cloud and locally, Chinese models will be everywhere.
Nah, I model hop constantly as I work with serving GLM and Kimi models and they're not nearly as good as Opus 4.5+ and GPT 5.2+ and it's not particularly close. They're good by standards set a generation or two ago, but they're really not competitive with where the frontier models are at now.
They compete with "mini" or "nano" model classes quite well given the price of inference. You'd need to "model hop" anyway, using Opus for everything is quite wasteful.
3 replies →
Guess it really depends on what you use them for. I've been able to built whole apps with them, not slop. Kimi is quite good at design, for 3D, I noticed Gemini 3.1 is excellent for basic to medium use cases.
I've tried both Opus and GPT 5.4, they also hallucinate just like the rest at a much higher cost.
The more you use a model overtime, the better you become with it. It's really hard to measure, my main metric lately has been tokens per second/time to complete task.
At this point I've the feeling frontier models are optimizing for benchmarks and one shot prompts.
If you actually use them you'll see that they are far from frontier models. They are much more cost-effective for what they are, but frontier they are not.
My understanding is that it's not that the _models_ are banned, but rather the _platform_ is banned. It is acceptable to host, say, `deepseek-r1-distill-qwen-7b` and run it yourself, for example. It is not acceptable (to the authors of these bans) to download the DeepSeek app and run it on your work device.
I just left a job for a German B2B software company which sold primarily to large automotive, defense, and aerospace companies. Several of our customers specifically banned anything with the word "DeepSeek" -- hosted or self-hosted.
There's still a lot of naivety on what the difference is between models and platforms, and its easier for a lot of these big companies to just make a blanket statement like "nothing DeepSeek" than for their procurement teams to try to understand and negotiate with each vendor. They don't see the potential benefit over the potential risk of somebody misinterpreting or getting it wrong, so they outright ban it.
Most people that approve or buy software simply also just don't understand how models are being trained or if it's possible/how far a model could go to "introduce backdoors." A backdoor could be, from a business perspective, a model which has been trained to give answers that could hurt western business in a "strict text mode" or produces payloads in a programmatic mode that are intentionally trained to introduce software vulnerabilities.
Anyone can make arguments against these for a variety of reasons (looking at the transparency of both sides and comparing, etc) but for many reasons today and for better or worse, many Chinese models are being banned on big software contracts, which gets back to the title of the article
Thing is these models can also be a propaganda machine whether you run it locally or not. This is true no matter the origins. Chinese LLMs will never shit-talk CCP, and it will always give a rosy depiction of the Chinese government. It's perfectly understandable if companies don't want things like that. US/EU models have these problems too, but at least there are some ways to fight that: with a lawsuit or a megaphone on social networks. With Chinese models there is nothing you can do.
You are sending all your prompts code and files there. So ofcourse its an issue
1 reply →
We aren't allowed to use any unauthorized models even locally.
> They are winning because West is forbidden to use Chinese models for anything work-related.
Because the models hosted in China are not trusted. This is 100% a part of what makes up commercialization.
Is anyone outside the US trusting anything hosted in today's US? If so, why?
I would say that both US and China are using the data we trust upon them for industrial espionage. So don't use their models if you are working defence or other sensitive areas
Deepseek is a fraction of the cost of western LLM and still just as good. I say it's also related.
Do we have any solid evidence these models can outperform Western models in terms of quality? Or is it more: because they are forbidden, they can't get enough training data, visibility etc. to compete?
Scroll down to the leaderboard - https://arcprize.org/leaderboard
Spoiler alert - they are all towards the bottom of the leaderboard. People come up with a wide variety of excuses for why they are not used despite being offered for significantly lower cost, but the answer is simply because they don't perform well enough for now.
There isn't even deepseek V4.
I'd rather trust LLM arena leaderboard, which puts it on par with sonnet.
1 reply →
You’re saying if we were allowed to use e.g. qwen more broadly the US wouldn’t be in the same strategic position? We have the best models…we own all the companies that make the best infra and the hyper scalers…I don’t think “oh we can use Qwen now?” Would exactly devastate the US
> I don’t think “oh we can use Qwen now?” Would exactly devastate the US
You'd be surprised how useful it can be to fine tune it in enterprise.
Well definitely but we have plenty of sanctioned OSS options for that
Qwen's open models are quite small compared to Kimi, GLM and DeepSeek Pro, which are often described as near-SOTA.
Why? So that even more American IP can pass through Chinese servers? Or because their near frontier models are heavily government subsidized?
>No, they are not. They are winning
You agree they are winning though, right? China is known for not playing fair, stealing industrial secrets, etc... that reputation matters and it's a good reason why the US is winning. Is the US perfect? No. Does the US play fair? No. Spare me the whataboutism in the comments. The bottom line is most people think the US is a safer bet and that's why we're winning. I personally wouldn't trust either government, but if I had to choose, I feel like I at least have a chance at secrecy and due process with the US. Obviously that is being eroded day by day, but you literally have no due process in China.