When will large Chinese models like Zhidu catch up to the level of Fable? Musk: Perhaps in the first quarter of 2027.

On June 18, a netizen asked on social media: "When will the Chinese large-scale model reach the level of Anthropic's Fable? The GLM-5.2 from Zhitu undoubtedly narrowed the gap." Tesla CEO Musk replied, "Possibly in the first quarter." On June 15, Zhitu announced the official launch and open sourcing of the new generation flagship large-scale model GLM-5.2. Unlike previous models that focused on real-time question-answering capabilities, GLM-5.2 focuses on "long-term tasks" - allowing AI to work continuously for hours, autonomously completing a large project like a human. In the front-end development evaluation system Code Arena with millions of global users participating in blind testing, this model ranks first among available models worldwide. Some AI industry insiders believe that Musk's prediction is relatively conservative, and the time difference in the level of models between China and the US may be less than 7 months. Demis Hassabis, CEO of Google's DeepMind, has also previously stated that Chinese AI models may "only be a few months behind" in terms of capabilities compared to overseas.