Foreign media: The threat of Anthropic Mythos prompted DeepSeek to launch its first huge financing.

According to DONGCHA Beating monitoring, The Information, citing people familiar with the matter, reported that DeepSeek has decided to abandon a founder-funded model and launch its first round of financing of over 500 billion yuan (about 7.4 billion USD). The direct trigger is the security model Mythos that Anthropic demonstrated in April. Mythos has the ability to find and exploit software vulnerabilities.

DeepSeek’s R&D team assessed that, when faced with models that require ultra-large-scale computing power and stacked data—and even models with destructive power at the national security level—relying solely on Liang Wenfeng’s personal wealth would not be able to compete in the computing arms race. It must introduce external capital to expand its computing power pool. Last year, to respond to U.S. export controls, DeepSeek tried to migrate its training and deployment systems from Nvidia’s CUDA ecosystem to Huawei chips. Rewriting the underlying software led to a delay of up to 15 months in the release of its next-generation model.

DeepSeek has previously maintained an extremely flat operating structure, with no human resources or public relations departments; everyone reported directly to Liang Wenfeng. With the completion of the massive financing, the team is now beginning to recruit functional roles such as legal, finance, and human resources.