A Beijing-based startup just released the world’s largest publicly available open-weight AI model, and the reaction from Washington’s tech orbit was immediate. David Sacks, a member of the President’s Council of Advisors on Science and Technology, took to social media to warn that America risks losing the AI race to China because of its own regulatory choices.
The catalyst: Moonshot AI’s Kimi K3 model, launched on July 16, 2026, packing 2.8 trillion parameters with native vision capability. It promptly climbed to the top of Arena.ai’s Frontend Code leaderboard with an Elo score of 1679, outperforming established Western models including Anthropic’s Fable 5.
The price gap that should worry Silicon Valley
Kimi K3 is priced at $3 per million tokens. Anthropic’s Fable 5 costs $15 for the same volume. That’s approximately 40% cheaper according to the research, though the dollar figures themselves represent an 80% difference.
The full weights of Kimi K3 are expected to be publicly released by July 27, 2026, meaning the open-source community will be able to build on, fine-tune, and deploy the model without licensing constraints.
Sacks sounds the alarm on regulation
On July 17, the day after Kimi K3’s release, Sacks posted his criticism of current US AI policy. Specifically, Sacks targeted bans on new data centers and proposed pre-approval requirements for AI models, arguing these policies amount to America “tying itself in knots” while Chinese competitors sprint ahead. His warning that America would “lose the AI race” came in his capacity as a member of the President’s Council of Advisors on Science and Technology and former AI and crypto advisor for the Trump administration.