OpenAI is considering a significant price reduction, lowering token fees.

```

OpenAI is brewing plans to implement significant price cuts for its AI services, directly targeting competitor Anthropic. A price war centered on enterprise customers is taking shape.

According to a report by The Wall Street Journal on Wednesday, people familiar with the matter revealed that OpenAI is evaluating a plan to drastically reduce token charges. Tokens are the basic unit for AI companies to measure and charge, and this move aims to seize Anthropic's market share, while also proactively responding to Anthropic's anticipated similar price cuts.

The discussions are still ongoing and subject to change. OpenAI CEO Sam Altman recently admitted at an event that the cost of AI use has become "a huge problem" and stated, "We will have many ways to help users get more value for less spend."

The price cut discussions bring financial pressure to both companies. OpenAI and Anthropic are both currently losing billions of dollars due to the enormous computing power needed for their AI systems to process queries and execute tasks, and substantial price cuts may further erode their already pressured profit margins.

Anthropic's rise, OpenAI under pressure

This price cut consideration by OpenAI reflects the competitive pressure it faces in the battle for enterprise clients.

Anthropic’s programming tool Claude Code has quickly become popular among software engineers, dramatically boosting the company’s revenue. This five-year-old startup’s valuation recently surpassed OpenAI’s for the first time. In response, OpenAI has made its own programming tool, Codex, a key focus of the company’s development.

However, enterprise enthusiasm for AI spending is cooling down. Some companies that have invested heavily in Anthropic’s products are now starting to seek more control over their expenses. Earlier this year, a senior executive at Uber stated that the company’s self-use AI budget for 2026 has already been used up; another company leader said last month that it’s difficult to directly link the improvement in AI programming efficiency to the rollout of new customer features.

These comments have sparked widespread discussion in Silicon Valley around "tokenmaxxing"—whether the practice of maximizing token usage to boost productivity without regard to return is reasonable.

Price war tests business models, adds uncertainty to IPO prospects

A price war will serve as an early stress test for both companies' business models, as they are both on the brink of highly anticipated IPOs. OpenAI and Anthropic currently account for the lion’s share of emerging AI product revenue, underpinning their rapid growth. But a long-identified risk for investors is that products from both companies are highly substitutable, making it relatively easy for customers to switch between them.

OpenAI this week has secretly submitted its IPO application, following in Anthropic's footsteps. Sam Altman stated in a recent Slack message to staff that the company plans "to go public within the next year."

OpenAI stated in its confidential filing that "some things are more convenient to do as a private company," but gave no further details. Against this backdrop, whether the price-cutting strategy can effectively increase market share without further widening losses will be a key issue for investors when assessing its IPO value.

Risk warning and disclaimerMarkets carry risks, and investments need caution. This article does not constitute personal investment advice and does not take into account individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions herein are suitable for their individual circumstances. If you invest based on this information, you do so at your own risk. ```