AI currently comes primarily from the USA, but competition in China is rising rapidly. Two companies now show that they can compete easily with the USA in AI. This is an unpleasant defeat for the USA. Additionally, the USA has temporarily banned “its” AI tools for other countries. The Chinese companies are not familiar with such bans.
When it comes to artificial intelligence, several large US companies, in particular, are in the spotlight: companies like OpenAI (ChatGPT), Anthropic (Claude), and also Google (Gemini) are likely known to most who have dealt with AI. This also shows the current dominance of US companies when it comes to artificial intelligence.
But that has been changing for some time. China’s leading AI companies are increasing pressure on Silicon Valley, and the USA has suffered a defeat against China. This is reported by TheVerge magazine.
Two Chinese AI models are said to be not only faster but also cheaper than US models
Which models are being discussed? Specifically, it concerns the two Chinese companies Moonshot AI and Alibaba. Both have introduced new models that they claim can compete with the best models from OpenAI and Anthropic. And the special thing about it: Only a fraction of the cost. ChatGPT has been struggling with the cost factor for some time:
- Moonshot AI presented the AI model “Kimi K3” on X.com. Moonshot stated that the model consistently performs better in its tests than almost any US system, with only OpenAI’s “GPT-5.6 Sol” and Anthropic’s “Claude Fable 5” lagging behind
- Alibaba, on the other hand, introduced its new model “Qwen 3.82” on X.com. According to the company, its performance is only surpassed by Claude Fable 5
The special thing about it is that the USA wants to actively restrict the use of particularly powerful models or has already done so: Claude Fable was only allowed to be used by a handful of designated US companies; the ban has now been eased and the model is available again in Europe, reports Bayerischer Rundfunk.
In China, they are taking the opposite approach: instead of hiding their most advanced models behind closed doors, they make them available to the public. This means that every user can freely download, modify, and further develop the models without fearing penalties or restrictions.
What is still a problem? So far, the evaluations only come from the companies themselves; there are currently no independent test reports on the performance of the models. Therefore, it is not possible to accurately assess how powerful the new models really are.
Both manufacturers, Alibaba and Moonshot AI, have already stated that they want to show the internal data or their training specifications.
Chinese models on the rise: And this is not the first time that Chinese AI models have put pressure on the American market or at least created a sense of urgency: In January 2025, the AI DeepSeek made headlines because it cost only a fraction of ChatGPT and is still considered powerful. And this is not the only company that has set ambitious goals:
The Chinese company Baidu operates the search engine of the same name in China and recently also introduced two new AI models: ERNIE 4.5 and ERNIE X1. These are also said to be more powerful but definitely cheaper than the competition from the USA: ChatGPT costs 700,000 euros per day: A new AI is said to not only be faster but also cost only a fraction at 7,000 euros
Your opinion is important to us!
Do you like the article? Then let us know!