China's GLM 5.2 AI model rivals top US labs
A new open-weight Chinese AI model is matching America's best for a fraction of the cost, which is bad news for Elon Musk and the big US AI companies.
- China's GLM 5.2 model now scores near the top US models from Anthropic and OpenAI on coding tasks, and ranks above Claude Opus on one popular leaderboard.
- The catch is price: testers say it does similar work at roughly one-sixth the cost, because anyone can download and run it for free.
- This threatens the huge bets US firms have made, since companies are starting to question whether expensive American AI is worth it.
- Musk's own Grok model is rated as one of the weaker ones, and his Tesla is fading as the US car market is forecast to shrink toward 2040.
- Cheap Chinese rivals are spreading globally — BYD cars sell for 20-30% less and are entering markets like Korea, even though most are blocked in the US.
Outlook: Cheaper Chinese AI could pressure US tech valuations and feed worries that the AI investment boom is a bubble.