China’s Z.ai Claims Its New AI Model Rivals Anthropic’s Mythos

Chinese artificial intelligence startup Z.ai has unveiled a new large language model, stating that its performance matches that of Anthropic’s Mythos system. The announcement positions Z.ai as a direct competitor in the global race for advanced AI. The company says its model, which it did not name in the initial release, achieves comparable results in reasoning, coding, and complex problem-solving tasks. According to Z.ai, internal benchmarks show the system is on par with Mythos, which is widely considered one of the most capable models currently available. “We are not just catching up; we are matching the frontier,” a Z.ai spokesperson said. The firm added that its model was trained with a focus on efficiency, requiring less computing power than rivals while delivering similar output quality. Z.ai’s claim comes amid intensifying competition between US and Chinese tech firms. While US companies like Anthropic and OpenAI have led the field, Chinese startups have been rapidly closing the gap, often with more limited resources due to export controls on advanced chips. Industry analysts note that independent verification of Z.ai’s claims is still pending. However, if confirmed, the model would mark a significant step for China’s AI sector, which has faced scrutiny over both technical capacity and data security. Z.ai has not yet announced a public release date or pricing for the new model. The company says it is currently in talks with enterprise clients for early access.