Back to Newsroom

OpenAI is gaining on Anthropic with business users, new data indicates

By Modelverse Editorial·August 20, 2026·2 min read
OpenAI is gaining on Anthropic with business users, new data indicates

Ramp’s expense‑management data shows that, as of July, Anthropic holds about 44% of paying business users among its 70,000 U.S. customers, while OpenAI has slipped to roughly 40%. Anthropic first overtook OpenAI in May with a 41% to 39% split, and OpenAI has not reclaimed the lead since. The data cover a broad mix of industries but are weighted toward technology firms that use Ramp’s corporate card and bill‑pay services.

In Q3 to date, OpenAI’s growth rate among this cohort exceeds Anthropic’s, according to Ramp economist Ara Kharazian, though a full quarter remains and the trend could reverse. Kharazian noted that developers increasingly favor OpenAI’s GPT‑5.6 Sol model, while Anthropic’s higher‑end Fable 5 tier has seen weaker adoption, partly due to its price and a 30‑day data‑retention requirement that drew criticism. Overall, the share of Ramp customers paying for any AI service rose from just over 50% in March to nearly 56% by July, indicating an expanding market despite the rivalry.

  • Anthropic ~44% market share, OpenAI ~40% (July)
  • OpenAI’s Q3 growth rate exceeds Anthropic’s among Ramp business users
  • Overall AI adoption among Ramp customers rose from >50% (Mar) to ~56% (Jul)
  • Model‑specific notes: developers favor GPT‑5.6 Sol; Fable 5 adoption hampered by price and 30‑day data‑retention rule

Why this matters

The shifting percentages suggest that enterprise AI spending is not yet locked to a single vendor, indicating that model performance, pricing, and data‑handling policies continue to drive short‑term preference swings rather than long‑term lock‑in.

Share this article

Found this insightful? Share it with your community on Reddit, X, or copy the link.

ai-newsbrieftechcrunch-ai

Footnotes & Primary References

Related content

Liquid AI Releases LFM2.5-DSpark Draft Models That Deliver Up to 3.18x Faster Decoding Without Changing Model Outputs

Three ~300M drafters bring speculative decoding to LFM2.5, delivering up to 3.18x faster decoding with identical greedy output. The post Liquid AI Releases LFM2.5-DSpark Draft Mode...

Read article

A third of web pages published since ChatGPT's launch show signs of AI authorship, study finds

ChatGPT and other AI models are now authoring and editing much of the new web.

Read article

Meet UPDF: A Lightweight Adobe Alternative Built for the Agentic Era

PDFs are easy to read and hard to change. AI can now summarize a 90-page contract in seconds, but it still won't rewrite the source file cleanly. UPDF is built for that second half...

Read article