OpenAI and Anthropic Race Toward IPOs as Their Own Business Model Draws Fire

The same technology that's pushed OpenAI and Anthropic toward trillion-dollar valuations is now the thing critics say could sink their public debuts.
Both companies are preparing for IPOs that would rank among the largest in tech history — Anthropic valued at $965 billion as of May, and OpenAI at $852 billion as of March — even as scrutiny mounts over the token-based pricing model underpinning their revenue.
What "Tokenmaxxing" Actually Means
Tokens are the billing units both companies use to charge for AI usage, whether through subscription tiers or pay-per-use API calls.
The aggressive, often unmonitored adoption of AI tools inside workplaces this year produced a new slang term, "tokenmaxxing," describing engineers using AI excessively under pressure to show integration, without regard for whether the spending made sense.
Uber, Microsoft, Salesforce and Meta have all since introduced spending controls to rein in employee AI usage after realizing the token-payment structure had become more expensive than it was worth.

The Criticism Getting Louder
Palantir CEO Alex Karp told CNBC last week that "something has gone completely wrong" with the token-payment model, as enterprise customers increasingly balk at the cost of pay-per-use consumption.
📰 Read Also: OpenAI Files for IPO at $850 Billion — Targeting September 2026 Debut
Tech critic Ed Zitron went further, arguing both companies' financials are too weak to support a public listing at all, comparing the situation directly to WeWork's collapsed 2019 IPO attempt.
Zitron specifically flagged that OpenAI burned $38.5 billion in 2025 against just $13.1 billion in revenue, calling Anthropic's reported $47 billion annualized revenue run-rate a "non-GAAP mish-mash" that leaned heavily on the temporary tokenmaxxing surge.
Why Customers Are Pulling Back
Some enterprise customers have already begun switching to cheaper, open-weight Chinese models like DeepSeek and Z.ai's GLM-5.2, which now ranks among the top 10 large language models globally while costing four to six times less than frontier US models.
📰 Read Also: Anthropic Files for IPO at $965 Billion — Beating OpenAI to Wall Street
Lindy CEO Flo Crivello said he moved 100% of his company's AI traffic to DeepSeek after dealing with "unsustainable" costs on Anthropic's Claude models, though he said he'd be open to switching back if prices come down.
Both OpenAI and Anthropic are reportedly weighing significant price cuts in response, a move that would ease customer pressure but further compress margins just as both companies court public investors.

How Anthropic and OpenAI Actually Compare
Anthropic's enterprise coding tool, Claude Code, has driven rapid revenue growth that briefly pushed the company's valuation past OpenAI's, and Anthropic expects to post its first adjusted operating profit this year.
📰 Read Also: SpaceX Passes Amazon in Global Valuation Race
OpenAI, by contrast, faces a steeper path to profitability, though it retains scale advantages as ChatGPT became the first app to reach 1 billion monthly users in May, roughly three years after its 2022 launch.
The SpaceX Precedent Looming Over Both
OpenAI's own advisers reportedly presented executives with a choice: wait until 2027 for a $1 trillion valuation, or list sooner at a lower valuation, with OpenAI leadership treating any reduction from the trillion-dollar target as a nonstarter.
SpaceX's own IPO debut this year, while historically large, has since cooled from its peak market cap, offering both AI labs a cautionary data point about how quickly investor enthusiasm can fade once a stock is actually trading.
💭 TheTrendsWire's Take
The gap between Karp's and Zitron's criticism and Anthropic's own optimism about approaching profitability is really a question of timing, not disagreement about the underlying problem: token-based pricing worked brilliantly during an unsustainable spending frenzy and is now being tested against customers who've actually started reading their bills. Whichever company goes public first will effectively become the market's referendum on whether AI revenue growth reflects real, durable demand or a temporary spike that's already correcting itself.
TL;DR
- OpenAI and Anthropic are preparing IPOs valued at $852 billion and $965 billion respectively.
- Palantir CEO Alex Karp says the token-payment model has "gone completely wrong."
- Critic Ed Zitron compared both companies' financial readiness to WeWork's failed 2019 IPO.
- Enterprise customers are switching to cheaper open-weight models like DeepSeek amid rising AI bills.
- Both companies are reportedly weighing price cuts that would further compress margins ahead of going public.
Read More
You might also like
Forever Stamp Price Rises Again This Weekend
Jul 7, 2026
SSA Retirement Benefit July Checks: Who Gets Paid Next
Jul 6, 2026
Dell Stock Rises After Trump’s Public Shoutout
Jul 6, 2026
The Five Studios Xbox Bragged About in 2018 Are Now Being Sold
Jul 6, 2026
Trump Accounts Go Live as NYSE and Nasdaq Ring Bells Together for the First Time
Jul 6, 2026
Buffett's Old Warning Is Back as Investors Split on an AI Bubble
Jul 5, 2026

Business & Finance Editor
Sarah Collins reports on markets, Wall Street, corporate news, and the global economy. She specializes in making financial news accessible to everyday readers.





