When Chinese AI company DeepSeek released its cheap and serviceable V3 model early last year, it sent shockwaves throughout Silicon Valley and beyond, roiling the stock market, shaking political confidence in American AI, and stoking new fears from the ever-churlish China hawks.
A year later, DeepSeek is preparing to launch its new V4 model — a development which could have major implications for US tech companies and the firms backing them.
According to a CNBC bulletin, DeepSeek’s latest version is “expected to be imminent” given the release-schedule of previous versions. Depending on how impressive V4 is when it hits, the AI-heavy Nasdaq could be in for a major upset, as could the tech companies listed on it.
Per CNBC, the Nasdaq composite fell 3 percent when DeepSeek V3 made its debut last year, and shares for the chip giant Nvidia plummeted 17 percent, wiping out $600 billion in a flash. While both recovered from the hits over time, it was a defining moment for DeepSeek, securing its reputation as a global player in the California-dominated AI space.
If the stock market gets its “part two moment” — a DeepSeek able to compete with current-gen models from Anthropic and OpenAI — things could get ugly. Amazon, Microsoft, Meta, and Google spent hundreds of billions of dollars on AI across 2025, and are expected to shell out another $650 billion in 2026.
Simply put, there’s a lot more money in the pot at this point, and even more being shuffled around based on those future spending forecasts.
Given that DeepSeek V3 was able to confound the American tech industry for a total reported production cost of under $6 million — using lower-powered Nvidia chips, no less — there’s only one thing the AI industry can do this week: buckle up.
More on Chinese AI: China Planning Crackdown on AI That Harms Mental Health of Users