Uploaded August 2026 | Updated September 2026, 3 weeks ago
The market wiped out $3.3 trillion on a price nobody checked.
How we're taking advantage of news like this is inside
skool.com/coinpicksgenesis, link in my bio, a dollar a month.
Semiconductor stocks have shed $3.3 trillion since June 22. A big part
of that was Kimi K3, a Chinese model from Moonshot that shipped July
16. Wall Street called it the next DeepSeek moment. China builds
something great, gives it away for pennies, American AI companies lose
pricing power.
Then you read the price list.
Moonshot charges $3 per million tokens in and $15 per million out.
That is the same list price as Claude Sonnet 5, an American model. It
is also roughly five times what Moonshot charged for their own
previous generation.
They did not undercut anybody. They raised their own prices to match
the West.
The model got bigger, not cheaper. 2.8 trillion parameters, and more
parameters means more electricity and more cost. For three years every
forecast assumed models get smarter and cheaper forever. This one got
smarter and way more expensive, which quietly broke the assumption
holding up one of the most bearish cases on AI.
The market did not get a fact wrong. It got the direction wrong.
Follow for the moves the news skips.
The market wiped out $3.3 trillion on a price nobody checked.
How we're taking advantage of news like this is inside
skool.com/coinpicksgenesis, link in my bio, a dollar a month.
Semiconductor stocks have shed $3.3 trillion since June 22. A big part
of that was Kimi K3, a Chinese model from Moonshot that shipped July
16. Wall Street called it the next DeepSeek moment. China builds
something great, gives it away for pennies, American AI companies lose
pricing power.
Then you read the price list.
Moonshot charges $3 per million tokens in and $15 per million out.
That is the same list price as Claude Sonnet 5, an American model. It
is also roughly five times what Moonshot charged for their own
previous generation.
They did not undercut anybody. They raised their own prices to match
the West.
The model got bigger, not cheaper. 2.8 trillion parameters, and more
parameters means more electricity and more cost. For three years every
forecast assumed models get smarter and cheaper forever. This one got
smarter and way more expensive, which quietly broke the assumption
holding up one of the most bearish cases on AI.
The market did not get a fact wrong. It got the direction wrong.
Follow for the moves the news skips.
