OpenAI is considering major reductions in what it charges for its artificial intelligence models. The move comes as the company braces for aggressive pricing from rival Anthropic. Sources familiar with the discussions told The Wall Street Journal that OpenAI expects Anthropic to slash its own rates soon. So the ChatGPT maker wants to strike first.
Tokens. That’s the unit both companies use to bill customers for AI usage. OpenAI weighs significant cuts here. The goal? Attract users drifting toward Claude, Anthropic’s fast-rising chatbot family. This isn’t subtle maneuvering. It’s a calculated bet on volume over margin in a market where compute costs remain punishing.
Anthropic has surged ahead in the valuation race. In late May it closed a $65 billion funding round that pushed its worth to $965 billion. That topped OpenAI’s most recent mark around $852 billion. The New York Times detailed the shift. The company also rolled out Claude Opus 4.8. It boasted better code generation and error detection. Momentum sits firmly with Anthropic.
But OpenAI still leads on revenue for now. It posted $5.7 billion in the first quarter of 2026. Anthropic trailed at roughly $4.8 billion. Yet the trajectory favors the challenger. CNBC reported that annualized figures show Anthropic closing the gap fast. Both firms lose money. Heavy infrastructure spending sees to that. The question becomes who can sustain the bleed longer.
Pressure builds from multiple fronts.
Google just cut its AI Plus subscription from $7.99 to $4.99 a month. It doubled the included storage to 400 gigabytes. The TechCrunch story framed it as a warning shot in the consumer AI wars. Chinese labs already offer frontier-level performance at fractions of the American price. One analysis showed Claude costing nearly nine times more than the cheapest alternatives for identical workloads. Developers notice. Some quietly test migrations.
OpenAI has focused more on business customers lately. It introduced models aimed at high-value professional work. The company even dropped some consumer experiments like the Sora video generator. PBS NewsHour captured the pivot. Anthropic built its reputation on enterprise reliability from the start. Claude’s safety features and consistent performance win converts in corporate settings. Price may decide the next wave of adoption.
Executives at both companies talk of coexistence. Sam Altman told CNBC he sees room for multiple winners. “I think there is a race to deliver the best technology and build the best business,” he said. Demand for AI, he added, will support several strong providers. Nice words. Market reality looks harsher. Microsoft, OpenAI’s biggest backer, now highlights its cheaper models as an edge over Anthropic. Mustafa Suleyman, Microsoft AI CEO, called Anthropic “extremely expensive.” Companies hunt value. They switch when the numbers favor it.
Recent pricing history adds tension. Both labs raised rates on their flagship models earlier this year. They ended generous enterprise flat-rate deals. Customers now pay full token prices. Some developers complain on forums. One Hacker News thread noted the shift away from subsidies. “They’re clearly not trying to offer the lowest-price-possible to drum up demand any more,” a commenter observed. Yet the latest signals point to reversal. Anticipation of Anthropic’s cuts has OpenAI ready to move.
Profitability remains distant for both. OpenAI projects losses of $14 billion this year in some estimates. Anthropic may post its first profitable quarter soon, according to leaks. That edge matters as both eye public markets. Anthropic reportedly leads the IPO race. A cheaper offering could help OpenAI defend its user base and revenue trajectory before either company lists.
The token economics get complicated fast. Prompt caching delivers discounts on repeated inputs. Anthropic offers steeper savings here at 90% versus OpenAI’s 50%. Context windows differ. Newer models add features like effort controls and dynamic workflows. Performance gaps narrow. When capabilities converge, price becomes the deciding factor. And. The market appears headed there soon.
Analysts watch enterprise seats. Heavy users once enjoyed discounts. No longer. A company spending hundreds per seat monthly feels the change. If OpenAI drops token prices sharply, it could reset expectations across the industry. Google already pressures the low end. Chinese competitors squeeze the middle. The frontier labs risk margin collapse if they follow too aggressively.
But staying expensive carries its own risks. Developers cite performance first. Value follows close behind. One forum post warned OpenAI could lose ground if Chinese labs match GPT-level results at a fraction of the cost. History in software favors the pragmatic buyer. Loyalty lasts only as long as the numbers work.
OpenAI’s deliberations remain internal. No official announcement has come. The WSJ report, however, carries weight. Sources describe the discussions as active. They tie directly to expected Anthropic action. A price war. That’s the phrase used. It may arrive sooner than many predicted.
Consumers could benefit. Businesses too. Lower barriers mean broader experimentation. Yet the companies must still pay for GPUs, power, and data centers. Efficiency gains help. They rarely keep pace with demand. The winners will balance aggressive pricing with operational discipline. So far neither has cracked sustainable profits at scale.
The AI race has entered a new phase. Valuation. Revenue. Pricing. All three metrics now move in tight formation. Anthropic’s recent funding win gave it breathing room. OpenAI responds with potential price aggression. Watch the token rates. They will signal who blinks first in the battle for users.
WebProNews is an iEntry Publication