Enterprise AI vendors face pricing squeeze

OpenAI and Anthropic are now being judged on more than top-line growth. Reported first-quarter revenue figures published this week put OpenAI at about $5.7 billion and Anthropic at about $4.8 billion, preserving OpenAI’s lead for the quarter. But the more important split in the market is emerging lower down the income statement: pricing, customer mix and whether enterprise demand holds up if token prices keep falling or usage keeps rising. ### Why does a revenue lead not settle the race? The Information reported on May 21 that OpenAI generated about $5.7 billion in first-quarter revenue, nearly $1 billion more than Anthropic, citing people with knowledge of the financials. The same report said OpenAI’s growth was helped by Codex, business sales and ad testing in ChatGPT. CNBC reported on May 20 that Anthropic generated $4.8 billion in the first quarter and told investors it expected $10.9 billion in second-quarter revenue. (theinformation.com) CNBC said that target, if reached, would give Anthropic its first profitable quarter, according to a source familiar with the matter. Those figures point to two different commercial stories. OpenAI still appears larger by quarterly revenue, while Anthropic appears to be leaning harder on enterprise demand and usage-heavy coding products to improve margins faster than rivals, according to CNBC and multiple follow-on reports. (theinformation.com) ### Where is the pricing squeeze coming from? Investing.com described the pressure as a “token pricing crisis,” arguing that enterprise AI vendors face a market where model access is becoming cheaper even as compute costs remain high. (cnbc.com) That dynamic matters because enterprise contracts have to absorb both falling unit prices and rising customer expectations for more usage, more reliability and more customization. CNBC reported in April that Anthropic had moved away from flat-rate enterprise pricing toward per-token billing, a change that better matches revenue to actual use. That shift suggests at least some vendors are trying to protect revenue quality rather than maximize headline adoption through broad fixed-price deals. Ropes & Gray said in a May quarter review that AI is pressuring seat-based software models and raising the bar for vendors that cannot tie pricing to workflow context, proprietary data or usage-based monetization. (investing.com) That is not a company filing, but it aligns with the broader move away from simple subscription logic in enterprise AI. ### Why does enterprise mix matter more now? (cnbc.com) Ramp said on May 13 that Anthropic passed OpenAI in paid business adoption for the first time in its AI Index. Ramp said Anthropic adoption rose to 34.4% of businesses in April, while OpenAI fell to 32.3%. Ramp’s vendor page still showed OpenAI with the broadest footprint inside the generative AI category, with 78% of organizations using a vendor in that category also using OpenAI as of May 2026. (ropesgray.com) Taken together, those two data points suggest OpenAI remains widely distributed while Anthropic is gaining share in paid business usage. Proactive Investors reported that Claude Code was generating $2.5 billion in annualized revenue, citing enterprise adoption as a key driver of Anthropic’s growth. (ramp.com) That figure should be treated as a media report rather than a company disclosure, but it fits the pattern described by CNBC and Ramp. ### What are buyers likely to watch in the next round? (ramp.com) May 2026 data from Ramp and future company disclosures are likely to give the next read on whether vendors can keep revenue growing without giving up too much on price. Buyers will also be watching whether per-token billing, coding agents and enterprise bundles produce steadier margins than broad consumer-style distribution. That conclusion is an inference from reported revenue, adoption and pricing trends. (proactiveinvestors.com) The next concrete milestone is Anthropic’s projected second quarter. CNBC reported the company told investors it expected $10.9 billion in revenue and its first profitable quarter, while OpenAI’s next disclosed growth markers are likely to come through further reporting on business sales, Codex adoption and any updated financial leaks or filings. (cnbc.com) (ramp.com)