💥🇨🇳👁 China's Kimi K3 AI just shot past Claude and GPT at 90% less cost

Moonshot AI released Kimi K3, an open-weight model that immediately topped the Frontend Code Arena leaderboard, jumping from 18th place to 1st in a single release

With 2.8 trillion parameters, K3 becomes the largest open-source model to date. In six of seven frontend domains—brand marketing, reference-based design, data analytics, consumer products, simulations, and content creation tools—K3 ranked first, losing only in Gaming to Claude Fable 5.

This breakthrough is part of a broader wave of Chinese developer achievements:

🟥 Chinese open-source models are now 60-90% cheaper than Anthropic and OpenAI on equivalent tasks. In under two years, token costs for GPT-4-level work dropped from $60 to roughly $0.50 per million

🟥Chinese open-weight models went from under 1% of global API traffic to over 30% in one year

🟥 A recent MIT study found Chinese open-source models have surpassed US models in total downloads, despite the fact that China spent roughly one-tenth of what the US did

🟥 Chinese AI firms have seen their valuations soar: Moonshot went from $4.3 billion to $30 billion in six months—a sevenfold increase.

While US labs keep their best models closed and charge premium access, China is giving away the weights and making its money on scale, adoption, and ecosystem lock-in. The same pattern that transformed manufacturing (copy, improve, commoditize) is now playing out in AI.

👍Boost us | Chat | @geopolitics_prime