OpenAI is weighing steep token price cuts to fend off Anthropic, escalating a price war that threatens margins at both companies ahead of their IPOs.
OpenAI is weighing steep token price cuts to fend off Anthropic, escalating a price war that threatens margins at both companies ahead of their IPOs.

OpenAI is weighing steep token price cuts to fend off Anthropic, escalating a price war that threatens margins at both companies ahead of their IPOs.
OpenAI is evaluating significant reductions to its token pricing — the core unit of AI billing — in a bid to reclaim enterprise market share from Anthropic, whose Claude Code tool has surged in popularity among software developers, according to people familiar with the matter.
"AI usage cost has become a huge problem," OpenAI Chief Executive Officer Sam Altman said at a recent event, adding that the company is exploring ways to help users "get more value for less spend."
The pricing review comes as OpenAI's Q1 2026 revenue reached nearly $6 billion, roughly $1 billion ahead of Anthropic, according to The Information. But both companies continue to burn billions on compute infrastructure — OpenAI projects it will not reach profitability until 2030 and expects to spend $600 billion on compute by that year, down from an earlier $1.4 trillion plan. Anthropic is closer to breakeven, nearing its first quarterly profit after locking in a $35 billion chip financing deal with Apollo and Blackstone.
The price war represents an early stress test of both companies' business models just as they prepare to go public. OpenAI confidentially filed its S-1 on June 8, one week after Anthropic's own filing. With OpenAI targeting a valuation of up to $1 trillion and Anthropic already trading at $1 trillion on secondary markets — a 123% year-to-date surge — investors will scrutinize whether market share gains from aggressive pricing can offset the margin compression.
The competitive dynamics have shifted sharply in recent months. Anthropic's Claude Code has become the fastest-growing AI coding product, helping the five-year-old startup surpass OpenAI in valuation for the first time. In response, OpenAI has elevated its own coding tool, Codex, to a top corporate priority. Codex now counts 4 million weekly active developers, a 6.7x increase from roughly 600,000 at the start of 2026.
But enterprise appetite for AI spending is showing signs of saturation. Uber's management told investors earlier this year that the company had exhausted its 2026 AI usage budget. Another corporate customer said last month it struggled to link AI coding efficiency gains to new client-facing features. The comments have fueled a broader Silicon Valley debate about "tokenmaxxing" — the practice of maximizing token consumption without clear return on investment.
The pricing review puts OpenAI in a delicate position. The company's confidential S-1 filing acknowledged that "there are things we want to do that are likely easier as a private company," without elaborating. A price war could deepen losses at a moment when public market investors are demanding clearer paths to profitability.
OpenAI's secondary stock has risen only 11.3% year-to-date, compared with Anthropic's 123% surge, reflecting what OpenVC CEO David Shapiro described as OpenAI having "already grown into a significant portion of its valuation." The divergence suggests investors see Anthropic as having more upside as it approaches profitability.
For developers and enterprise customers, the outcome of the pricing battle will determine whether AI costs continue to fall or stabilize. If OpenAI and Anthropic are forced to raise prices to satisfy public market margin expectations, the era of subsidized AI could end abruptly — a dynamic already visible in GitHub Copilot's recent switch to token-based billing, which triggered bill increases of as much as 25x for some users.
The convergence of the SpaceX, Anthropic, and OpenAI IPOs within months represents the most consequential test of investor appetite for high-growth technology stocks in a decade. How the AI pricing war resolves will shape not just the two companies' valuations, but the cost structure of the entire AI industry.
This article is for informational purposes only and does not constitute investment advice.