Good morning.
Simmy Adelman here with Behind the Markets.
I want to start this week by walking you through something that happened at the end of last week that I think a lot of investors missed.
While many are still focusing their attention on SpaceX falling below its IPO price and chip stocks selling off, China made two moves that will matter a lot more to your portfolio over the next decade than anything that happened on the Nasdaq last week.
Move One: China Just Gave Away Its Crown Jewel
On Thursday, a Beijing-based AI startup called Moonshot AI released Kimi K3.
2.8 trillion parameters.
The largest open-source AI model ever built.
And they're giving the full model weights away to the public on July 27th.
Anyone on earth can download it.
For free.
To understand why that matters, you need to understand what "open source" actually means in this context.
When a company releases an open-source model, they're not just sharing the product.
They're sharing the blueprint.
Anyone can take Kimi K3, modify it, build on top of it, deploy it in their own products, and use it however they want.
No license fees.
No subscriptions.
No dependency on an American company.
Now here's the part that should get every investor's attention.
Kimi K3 benchmarks neck-and-neck with the most powerful systems from Anthropic and OpenAI on some tasks.
To be accurate — it still trails Claude Fable 5 and GPT 5.6 Sol on overall performance.
But on coding, general agents, and front-end development, it ranked first — above the leading US proprietary systems.
Bank of America put it plainly: despite persistent hardware and compute constraints from US chip export controls, K3 still delivered major capability gains.
China is innovating around our restrictions.
Not being stopped by them.
And the price?
Kimi K3 is priced at $3 per million input tokens.
Anthropic's comparable Opus 4.8 runs $5 per million input tokens, with their most advanced model, Fable 5, running $10.
OpenAI's flagship runs similar numbers.
China is offering frontier-level performance at a fraction of the price
Western companies are already noticing.
DoorDash, Siemens, and Airbnb have all started switching to Chinese AI models because they are cheaper, capable, and easier to run on their own infrastructure.
Move Two: Xi Takes the Stage
On Friday morning, Xi Jinping delivered the opening keynote at the World AI Conference in Shanghai.
It was his first in-person appearance at the conference since it launched in 2018.
His message was pointed.
"AI should not be a solo performance by a single country, but a symphony of international cooperation."
He wasn't talking about America.
He was talking to the rest of the world.
And the rest of the world was listening.
The day before Xi's speech, nearly 30 countries — including Russia, Brazil, and Indonesia — signed up to China's new World Artificial Intelligence Cooperation Organization, known as WAICO.
This organization gives Beijing direct influence over international AI standard-setting.
Think about what that means.
While America has been focused on export controls and chip restrictions, China has been quietly building a coalition of countries that will follow its lead on how AI gets governed, regulated, and deployed globally.
Xi also announced that China will provide 5,000 AI training and development opportunities to developing nations over the next five years, and will build international AI cooperation centers across Africa, Latin America, Asia, and the BRICS countries.
That is a Marshall Plan for AI in the developing world.
And China is writing the check.
The Part Nobody Is Talking About
Here is what ties all of this together.
The Trump administration briefly imposed export controls on Anthropic's most advanced models last month — Fable and Mythos — due to security concerns.
Then backtracked after backlash from Silicon Valley.
That episode told China something important.
America is nervous.
And nervous countries make reactive decisions.
China is making strategic ones.
Stanford University's 2026 AI Index found that the performance gap between the leading US and Chinese models has narrowed to just 2.7 percentage points.
Two point seven.
That is not a comfortable lead.
And while America debates export controls and chip restrictions, China is giving its best models away for free to every developing nation on earth and building an international organization to govern AI on its own terms.
What This Means for You
We’ve been saying for a long time now that the AI race is also a national security race.
This week proved it.
The companies that win this race won't just be the ones with the best models.
They'll be the ones with the most users, the most allies, and the most influence over how this technology gets built and governed.
China just made three major moves in that direction in 48 hours.
The question for investors is simple.
If China wins the open-source AI war the way it won the drone manufacturing war — flooding the market with cheap, capable technology until nobody else can compete — what does that mean for the American companies that have been charging premium prices for premium models?
And what does it mean for the infrastructure underneath all of it — the chips, the energy, the data centers — that still runs on American technology regardless of who builds the models on top?
Those are the questions we’ve been thinking a lot about.
Because here’s the thing most people miss.
Whether China wins the model war or America does, whether Kimi K3 takes market share from Claude or it doesn't — every single chip that runs any of these models is built on the same blueprint.
It doesn't matter whose flag is on the data center.
The royalty gets collected either way.
One quiet company owns that blueprint.
Dylan put together a full presentation on exactly who it is, and why he thinks this stock belongs in every investor’s portfolio.
If you haven't seen it yet, this is the week to watch it.
Have a wonderful Monday.
I'll see you tomorrow.
All the best,
Simmy Adelman, Editor-in-Chief
Behind the Markets
Found this helpful? Share it with others.
Written by Simcha Adelman