OpenAI and Anthropic, two of the most prominent names in artificial intelligence, are facing a double squeeze that threatens their profitability and the lofty valuations tied to their potential public offerings. Competition from cheaper AI models — many built by smaller startups or released as open source — is chipping away at their pricing power. At the same time, the cost of training and running frontier-level AI systems keeps climbing.
The price war nobody asked for
A wave of low-cost AI models has hit the market over the past year. Some are built by well-funded competitors, others by hobbyist communities. What they share is a price tag that undercuts the rates OpenAI and Anthropic charge for their flagship products. Companies that once defaulted to paying for premium APIs are now testing cheaper alternatives, and some have switched entirely.
OpenAI and Anthropic have built their business models around selling access to large, powerful models. The arrival of viable substitutes — some nearly as capable but far cheaper — is eroding the margin they can command. Neither company has disclosed exactly how much market share they have lost, but investors tracking the space say the trend is unmistakable.
The hidden cost of cutting-edge AI
Developing a state-of-the-art large language model is brutally expensive. Data center time, specialized chips, electricity, and the salaries of top researchers all add up. Both OpenAI and Anthropic have spent hundreds of millions of dollars — possibly billions — training their latest models. And the bill is only growing.
Operational costs are rising just as revenue growth faces new headwinds. The companies have responded by raising prices for some customers and by pushing subscription tiers, but those moves can only compensate so much. If the cost of serving a single API request continues to rise while competitors offer comparable results at a fraction of the price, the math gets ugly fast.
IPO valuations on shaky ground
Both companies have been valued in the tens of billions in private fundraising rounds. OpenAI was reportedly in talks for a valuation north of $80 billion. Anthropic has been valued at around $18 billion. Those numbers assume that a path to sustained profitability exists — and that the IPO market will reward them with even higher multiples.
The threat from low-cost models and rising costs puts that assumption in question. If profit margins shrink, the valuations that early investors have been banking on may not materialize. An IPO would still be possible, but at a lower price — or on terms that give less to founders and more to underwriters.
Neither company has made formal IPO filings yet. Their plans remain in the realm of speculation, but the window for a high-priced debut is narrowing as the competitive landscape shifts.
What investors are watching now
The next few quarters will be critical. OpenAI and Anthropic are expected to release updated financial metrics, and investors will be looking for signs that they can defend their margins. If revenue growth slows or customer churn accelerates, the IPO timeline may slip further.
Meanwhile, the low-cost AI movement shows no signs of slowing. New model releases appear almost weekly, and many are free or nearly free to use. The two companies that once defined the frontier are now fighting to prove they can still make money at the frontier — and that their brand and performance are worth the premium.
