BlockBeats report, on June 22, Jefferies stated that Zhipu's GLM-5.2 has entered the top three globally among large models, marking a significant milestone in China's AI development; however, U.S. restrictions on access to Anthropic's top models are unlikely to bring Zhipu significant revenue growth.
In a Chinese tech industry report released on June 17, Jefferies stated that GLM-5.2 ranks third on the Global Model Intelligence Index, behind only Anthropic and OpenAI—marking the first time a Chinese model has entered the global top three. The report noted that GLM-5.2 excels in programming and long-term agent workflows and has become one of China’s leading large models.
Zhipu released GLM-5.2 via Coding Plan on June 13 and opened its API on June 17. The model features a 744-billion-parameter MoE architecture with approximately 40 billion active parameters, supports a 1 million token context window, and has its weights open-sourced under the MIT license. Jefferies noted that, according to the Artificial Analysis Model Intelligence Index, GLM-5.2 ranks third globally and first in China; it also ranks fourth globally and first in China for coding, and second globally and first in China for agentic performance. On the Arena.ai frontend programming leaderboard, GLM-5.2 Max ranks second, behind only Fable 5 High, and above Claude Opus 4.7 and 4.8.
The report also noted that restrictions imposed by the United States on access to Anthropic's Claude Mythos 5 and Fable 5 may be difficult to sustain long-term. After Anthropic released the models on June 9, the U.S. government, citing national security concerns, requested that it restrict access by foreign nationals, including non-U.S. citizens working within the United States. Due to the short-term difficulty in distinguishing user identities, Anthropic temporarily suspended global access and strengthened identity verification measures.
Jefferies believes that the arrangement of "access to frontier models limited to Americans" is difficult to enforce and could harm the U.S. AI ecosystem. Since a significant number of AI researchers and engineers in the U.S. are foreign-born or non-U.S. citizens, excluding them from frontier models could slow research progress and even drive some talent and demand overseas.
However, the report suggests that the Fable 5 restrictions are unlikely to significantly translate into revenue for Chinese model companies. Jefferies channel surveys indicate that most users will not directly switch from Claude to GLM-5.2. GLM-5.2’s programming capabilities are considered close to Claude Opus 4.7 but still lag behind Opus 4.8. Even if developers adopt Chinese open-source models, they may use them via local deployment, cloud providers, or inference platforms like OpenRouter, without necessarily generating revenue for the model developers.
The report also states that developers typically use a mix of models for different tasks, including code generation, debugging, and long-context tasks, rather than fully migrating to a single model. This limits GLM-5.2’s ability to commercialize potential traffic spillover from Anthropic.
Jefferies noted that Zhipu still faces insufficient high-end reasoning compute capacity, particularly as long-context and agent workflows consume more resources, potentially impacting the delivery of enterprise-grade requirements. The report also highlighted intense price competition in China’s large model market: while GLM-5.2 offers approximately a 26% performance improvement over GLM-5.1, API pricing remains unchanged, indicating that model providers lack significant pricing power.
In addition, U.S. model vendors' anti-distillation technologies and export controls also pose long-term risks. Jefferies stated that these measures could restrict China’s ability to catch up with cutting-edge models and lead to a renewed widening of the performance gap between U.S. and Chinese models.
The core conclusion of the report is that GLM-5.2 demonstrates that Chinese models have entered the global top tier, but they have not yet truly “worn Anthropic’s shoes.” Under constraints such as revenue conversion, computing power supply, price competition, and U.S. technology restrictions, Zhipu is unlikely to achieve meaningful commercial gains from Anthropic’s access limitations in the short term.