DeepSeek Just Dollar-Store'd OpenAI's Entire Business Model
DeepSeek-V3 cost $5.6M. GPT-4 cost $100M+. China's budget AI labs aren't catching up—they're playing a different game entirely. Silicon Valley's compute moat was tissue paper all along.
DeepSeek-V3 cost $5.6M. GPT-4 cost $100M+. China's budget AI labs aren't catching up—they're playing a different game entirely. Silicon Valley's compute moat was tissue paper all along.
A $6M Chinese AI model just matched GPT-4o and Claude 3.5 Sonnet. NVIDIA lost $593B in a day. The AI moat was always a myth.
The AI lab that wiped $1T from Nvidia's valuation is now building its own silicon. U.S. export controls accelerated exactly what they were designed to prevent. Jensen should be worried.

DeepSeek-V3 at $0.14/M tokens vs GPT-4o at $2.50. The AI price war isn't coming — it's here, and U.S. companies are quietly switching teams.

DeepSeek's efficiency-first approach is undermining Nvidia's core pitch: that you need ever-more GPUs. With R1 trained for $5.6M on export-controlled H800s, the Blackwell sales deck just got significantly harder to close.

DeepSeek, Qwen, and Kimi match GPT-4 and Claude for pennies. Export controls didn't slow China — they made it leaner, meaner, and open-source.

DeepSeek R1 cost $5.6M to train and matched OpenAI's o1. NVIDIA lost $600B in a day. The open-source AI revolution is here and Wall Street's late.

OpenAI and Anthropic are getting squeezed as Google, DeepSeek, and Meta slash AI prices toward zero. The foundation model business is commoditizing faster than the valuations can handle.

DeepSeek trained a GPT-4-class model for $5.6M on nerfed chips and open-sourced the whole thing. The AI hype machine just met its reality check.