OpenAI has cut the price tag it is asking for in its latest funding round to $1.25 trillion. The Financial Times reports it first pitched $1.4 trillion, then lowered the number a week before the deadline as tech stocks fell. OpenAI has raised about $11 billion of the $30 billion it first sought. It now aims for up to $20 billion.
Why the price dropped
Investor Rory O'Driscoll says this is a simple pricing question, not proof the AI boom is cracking. Two companies are about to ask investors for money. One is bigger, growing faster, profitable and about to go public, with shares expected to trade at $1.5 to $2 trillion. The other is smaller, growing slower and still losing money, and its shares are hard to sell. The second should price lower.
His warning is about demand. If people keep using AI, the chips and data centers get used and the bills get paid. If demand cracks, the firms with huge cash piles are hurt last. Smaller data center companies such as CoreWeave are hurt first.
Signs of a nervous market
Alphabet lost about $500 billion in market value in what was reportedly its worst day in about a decade. Google Cloud revenue jumped 103% to $30.7 billion. But Search growth slowed to 12%, its slowest pace since 2023. Free cash flow was negative $9.07 billion for the second quarter in a row, as spending of $51.36 billion outran the cash coming in.
Shopify fell 15% during the recording of the podcast that covered all this, after coming in slightly short of expectations. Any missed number is being punished.
Anthropic's slowing growth
Anthropic's filing shows revenue of about $5 billion, then $11 billion, then $17 billion over three quarters. Growth from one quarter to the next fell from about 100% to about 50%. Over a full year, that is roughly 12x this year and about 2x next year at the same pace. Rory called it a slowdown to numbers that are still magnificent.
Jason Lemkin of SaaStr reads it as pricing pressure, not weak demand. He says OpenAI is selling roughly equivalent products at 30% off, and freely available models add pressure. The pricing pressure may cut Anthropic's ability to charge a premium.
What it means for business owners
Lemkin moved SaaStr Connect, an AI matching app, to Haiku 5.5 and a cheaper mix of models. Total costs fell 72%. The work ran faster and at higher quality. He says nothing he had run through AI agents, which are tools that act on their own, had gotten cheaper per use since early 2025.
- Check what your AI tools cost per task, not just per month.
- Test cheaper models on routine work.
- Compare quality as well as price before switching.
One founder's result is not a rule. But the price of AI work may be falling for the first time in a while.
The takeawayAI prices are under pressure, so test cheaper models on routine work before assuming your current costs are fixed.