Moonshot AI, a Chinese startup, has released Kimi K3, a new open-weight AI model it claims rivals offerings from OpenAI and Anthropic at a much lower cost. The release has drawn comparisons to the DeepSeek moment from early 2025, which also shook markets when a low-cost Chinese AI model went public.
Analysts at Stifel say the launch could be a positive for cybersecurity and AI infrastructure companies. The firms they name include CrowdStrike, Datadog, JFrog, MongoDB, Cloudflare, Palo Alto Networks, Snowflake, and Twilio.
CrowdStrike Holdings, Inc., CRWD
The logic follows what’s known as Jevons Paradox. When something gets cheaper, people tend to use more of it. Lower AI costs could drive higher consumption, which would benefit the infrastructure and security software vendors that support all that usage.
Stifel analysts, led by Brad Reback, noted that cheaper AI applications could give AI-native companies more room to spend on research and distribution. That could make them stronger competitors over time.
Reback pointed out that lower token costs tend to help software vendors’ profit margins. Less spending on AI features means better unit economics. Security vendors, in particular, are seen as durable because more AI usage tends to mean more vulnerabilities to defend against.
The application layer, Reback said, remains the most uncertain part of the stack.
Moonshot AI was founded by Yang Zhilin, a former researcher at Google and Meta. The company is currently raising funding at a valuation of around $30 billion. Alibaba and Tencent are among its backers.
The Trump administration is currently looking at whether to restrict the use of Chinese AI models in the United States.
Kimi K3’s specs — 2.8 trillion parameters and a one-million-token context window — require far more memory than earlier AI models. That points to strong demand for memory chips, a market dominated by SK Hynix and Samsung Electronics.
Stanley Tang of Sumitomo Mitsui DS Asset Management told Bloomberg that memory makers are well-positioned due to the limited number of suppliers. He said demand is unlikely to fall if models like Kimi become widely used.
Gary Tan of Allspring Global Investments agreed, saying China’s open-source AI push will require more computing resources, driving demand for networking and memory hardware.
Nvidia and AMD are in a less clear position. Both face renewed questions about their premium valuations as cheaper AI alternatives keep arriving. Alibaba and MiniMax both launched new models in the wake of Kimi K3’s release.
The post Kimi K3: How China’s New AI Model Could Boost Cybersecurity Stocks and Memory Chips appeared first on CoinCentral.
