There are basically two tiers of "Chinese models" in this context, the "edge" sized ones with ~30B parameters or less, and the big ~1T models that can basically only run in the datacenter.

I don't think it's as simple as saying China's hosting is subsidized, they have generally cheaper electricity and labor costs than in the US and don't have access to the top tier models, and a large internal market where the big models are the best thing they can run with what they have. So obviously they max out on their top models (which are trained with their hardware market in mind, not ours) and get the economy of scale from that, and can run generally the same hardware for less money than in the US because

The edge models are very cheap to run and can do so on inexpensive hardware. They are like 95% cheaper to run than Haiku, so the math is in their favor for certain batch workloads. Most people just run the models for themselves when they do that without making it available on openrouter or whatever, because you can just provision a gpu node and use it as needed, and it's not that expensive to run this family of models.

Is your problem that you want to call Chinese models hosted in the US because you're worried about the data handling?

▲

pants2 3 hours ago | parent [-]

I obviously don't know the full economics of the Chinese-hosted models, but estimates[1] put the cost of hardware (servers + networking) at 70-80% of the total cost. Those things aren't meaningfully cheaper in China, so serving DeepSeek at 1/3 the cost of the cheapest US provider doesn't really compute unless it's heavily subsidized or we believe that Chinese engineers are just that much better at optimization.

Edge models, yes, they can be convenient to run batch jobs locally. I still would argue there's no economic benefit over paying for models. Haiku has a bad price/perf but others in that class are significantly cheaper in hosted APIs.

Doesn't matter what I think, the reality is that the majority of enterprises (where the real $ comes from) will not consider sending their data to China.

1. https://epoch.ai/data-insights/ai-datacenter-cost-breakdown

	▲	torginus 43 minutes ago \| parent [-]
		Hardware is arbitrarily priced, with the floor being as little money as it costs to make it, and the ceiling being how much competitors are willing to pay for it - the latter is much more of the driver of current pricing in the West than in China. In a free market, the country would not matter, but Chinese models are often running on domestic hardware which does not directly compete with Nvidia GPUs and thus they can't get away charging as much for it.