OpenAI could cut ChatGPT prices to win you over from Claude

The AI industry’s battle for customers could soon be good news for users’ wallets. OpenAI is reportedly preparing steep cuts to its token costs to claw back enterprise momentum lost to rival Anthropic.

OpenAI is considering reducing the cost of tokens, the units used to measure and bill for AI usage, before a possible similar action by its fast-rising competitor, The Wall Street Journal reported, citing people familiar with the matter.

The potential change comes as businesses increasingly push back against the soaring costs of deploying AI. Token costs can add up quickly for companies that use AI at scale. This is particularly true for enterprise customers running coding assistants, agents, and other productivity tools that use huge amounts of computing resources.

OpenAI’s pricing discussions also show how much pressure Anthropic has been able to put on the market. In recent months, the younger company has gained momentum, thanks to the popularity of its offering, Claude Code, among software developers. That success has apparently helped boost Anthropic’s revenue and even helped the startup eclipse OpenAI’s valuation at one point.

Since then, OpenAI has intensified its coding ambitions, placing greater internal emphasis on Codex.

But cheaper AI won’t automatically mean healthier profits. Both Anthropic and OpenAI are already spending billions on infrastructure to train models and serve users. More aggressive price cuts could squeeze margins even further.

Some corporate customers are also beginning to question if their AI spend is delivering enough bang for their buck. Those concerns have ignited wider conversations in Silicon Valley about “tokenmaxxing” — the notion of consuming as many AI tokens as possible to boost productivity, even when the financial return isn’t always obvious.

A pricing war could give some clues about how sticky these AI platforms truly are, and investors are closely watching. OpenAI and Anthropic are the big revenue winners in this AI boom, but they have a shared problem: Customers are more likely to churn than in many legacy software businesses.