OpenAI is offering some of Anthropic’s biggest customers discounts of up to 90% to move their work to its models, according to a report from The Information published on October 24. The Decoder’s account of the report says Replit moved most of its AI spending to OpenAI under a contract worth more than $100 million a year after a discount of more than 30%, while Rippling lifted OpenAI’s share of its coding work from 16% to 52% as Anthropic’s fell from 57% to 18%. DataCamp’s chief executive said OpenAI’s pricing cut his company’s costs by 80% to 90%. None of the companies has publicly confirmed the contract terms, and the reports include no statement from Anthropic.
The Deals Behind The Shift
According to The Information, whose reporting The Decoder and other outlets relayed, OpenAI’s discounting reaches beyond a single account. At Rippling, the HR software company, Anthropic’s models handled 57% of engineers’ coding work in May while OpenAI’s handled 16%. Those shares have now flipped: OpenAI handles 52% and Anthropic 18%. Rippling chief executive Parker Conrad said the company switched because OpenAI’s pricing was more competitive.
DataCamp shows the same pattern at a smaller scale. Its chief executive, Jonathan Cornelissen, said OpenAI offered a 40% discount on models used for internal product development, provided the company also used Codex. He estimated costs could fall by 80% to 90% if the offer extended to the products DataCamp sells. A US bank with more than $7 billion in assets reportedly had its ChatGPT seats doubled to 500 employees at no extra charge, paying under $7 per user instead of the usual $20 or more.
The Decoder adds that OpenAI is partly buying market share at the expense of its own margins. It has also cut its own costs through its cheaper Sol and Luna model lines. Both labs face growing price pressure from China, where models from DeepSeek and Kimi deliver comparable performance for a fraction of the cost.
Replit’s Switch, Now Priced
The largest single figure in the report belongs to Replit. OpenAI offered more than 30% off a contract worth over $100 million a year, after Anthropic declined to extend Replit’s earlier discount unless it committed to higher spending. Replit’s chief executive, Amjad Masad, had already announced the switch onstage at TechCrunch Disrupt on October 16, saying the company was no longer working with Anthropic and citing cost, partnership and performance. On performance he was blunt: “I think ChatGPT is better now.”
The split has proven messier than the headline, as Replit’s switch to OpenAI as its default already showed. A person familiar with the deal told TechCrunch the Anthropic partnership is still in place, and Replit’s own documentation still lists seven Claude models. Masad later softened his own line, saying Replit remains model agnostic and that OpenAI is now its preferred partner for powering Replit Agent, based on performance, economics and the strength of the partnership. Most users stick with the default, so the default captures the bulk of the spending even when every previous option stays available.
Why Anthropic Cannot Simply Match
Anthropic’s problem is that discounting is expensive when margins are already under pressure. Its growth still held at 45% despite slowing from the second to the third quarter, according to an investor memo cited in the reporting. Its computing costs in the second quarter were $5.6 billion, less than half of its $11.5 billion in revenue, while OpenAI has forecast that its computing spending this year will significantly exceed its revenue.
Anthropic also has a policy of ending discounts when customers reach their contract limits, which contributed to losing the Replit business. Some of its biggest customers, including Cognition, Meta and Microsoft, are developing their own technology or alternatives to reduce what they spend on Claude. Notably, no benchmark appears anywhere in the reporting on these switches: the reported trigger was a withdrawn discount, not a failed evaluation.
Why It Matters
OpenAI’s discounting turns model choice into a procurement decision. When two labs’ models are close enough at the application layer that a 30% price gap outweighs any remaining capability gap, the invoice decides where the work goes, and the default slot decides who collects it. That is why a coding platform’s default matters more than its menu: most users never change it.
The timing sharpens the stakes for both labs. OpenAI is lining up its planned $30 billion raise as a bridge to a possible 2027 listing, while Anthropic is preparing its own public debut and expanding its enterprise reach through Anthropic’s full platform on Google Cloud. Discounted volume bought now becomes the revenue story each lab tells public investors later.
Background
OpenAI and Anthropic have spent 2026 competing for the same enterprise coding workloads, with Replit once ranking among Anthropic’s top five customers before its October switch. Both labs are heading toward public markets while spending heavily on computing capacity, leaving enterprise discounts as the most visible weapon in a rivalry previously fought on benchmarks.
Source: The Decoder, AI Weekly
0 Comments