OpenAI price war pulls customers from Anthropic

OpenAI is cutting prices sharply to win large business customers away from Anthropic, and the tactic is working. According to The Information, the ChatGPT maker has offered steep discounts to companies that had been among Anthropic's biggest accounts, in a price war that is reshaping who powers corporate software development.

The report, published October 24 under the headline "Anthropic's Margin is OpenAI's Opportunity," draws on customer executives and a person familiar with the deals. Neither OpenAI nor Anthropic has commented publicly on the figures, and the discounts described are negotiated deals rather than published list prices. Even so, the named accounts and the scale of the shifts give the clearest picture yet of how price, rather than model capability alone, is deciding enterprise AI contracts.

The accounts that moved

The clearest example is Rippling, the human resources software company. Chief executive Parker Conrad said Anthropic's models handled 57 percent of the coding work done by the company's engineers in May, while OpenAI handled 16 percent. That split has since reversed: OpenAI now handles 52 percent and Anthropic 18 percent, Conrad said, because OpenAI proved cheaper.

Replit, the coding platform, made a larger move. According to the report, it shifted most of its AI spending to OpenAI in a contract worth more than $100 million a year, after OpenAI offered a discount of more than 30 percent. Anthropic had declined to extend a discount it had been giving Replit unless the company committed to spending more.

Replit chief executive Amjad Masad said in mid-October that the company was no longer working with Anthropic, citing cost, partnership and performance, and saying ChatGPT was better now. He later softened that position, describing OpenAI as the preferred partner for powering Replit Agent, while a person familiar with the relationship disputed that the split was complete. Replit's documentation still lists Claude models, though OpenAI is now the default.

DataCamp, the online learning company, is another account to move. Chief executive Jonathan Cornelissen said OpenAI's pricing cut his company's costs by 80 to 90 percent. A separate account of the deal put a discount on OpenAI's Codex coding agent at the centre of the arrangement, on condition DataCamp use that tool. A large American bank also doubled its ChatGPT seats to 500 employees at no extra charge, paying under $7 per user against a standard rate of $20 or more, according to the report.

Why the timing matters

The discount push lands at an awkward moment for Anthropic. The company is expected to list on public markets in mid-November, with reports putting its target valuation near $2 trillion and a possible capital raise on a similar scale. An investor memo cited in the reporting says Anthropic's growth slowed from the second quarter to the third quarter, which makes losing marquee accounts a difficult story to carry into a listing.

Both companies have been busy on the product front alongside the pricing fight. Anthropic recently merged its Claude chat and Cowork tools for enterprise users, while OpenAI used its developer event to push always-on agents for developers. Those launches show the contest is not only about price, even when price decides a given deal.

OpenAI is paying a price for the strategy too. The Decoder, summarising the reporting, said OpenAI is partly buying market share at the expense of its margins. The discounts sit alongside cheaper, more efficient model lines, including the Sol and Luna families, which the company says cost less to run. But giving away seats and cutting contract values by a third or more eats into the revenue each customer produces.

Both labs also face pressure from cheaper rivals. Chinese firms such as DeepSeek and Kimi now offer comparable performance for a fraction of the cost, which gives every large customer a credible threat to walk away. That pressure helps explain why a customer like Replit could treat leading models as close enough substitutes for price to decide the deal, even in coding, where model differences are supposed to matter most.

OpenAI, meanwhile, is reported to be raising tens of billions of dollars at a valuation above $1.4 trillion, in talks involving SoftBank and Abu Dhabi's MGX, according to Reuters reporting relayed by Dealroom. The raise is timed ahead of Anthropic's expected listing, putting the two labs in front of the same pool of investors on overlapping timelines. For customers, the contest has a practical payoff: the biggest AI contracts are now negotiated on price first, and the savings on offer run into the millions. Sources: SiliconReport, Techpresso, Dealroom.