Zvi newsletter analysis of Ramp AI Index confirms Anthropic's slight extended lead and power-law spending distribution
Ramp July data: Anthropic at 43.5% U.S. enterprise adoption vs. OpenAI's 39.7%
Ramp spend data provides a concrete, transaction-based measure of which AI providers have penetrated U.S. business workflows. The concentration of spending in a small fraction of firms shows most businesses remain at the margin of AI investment.
The full picture
Per Ramp's July spend data, Anthropic reached 43.5% of eligible U.S. businesses vs. OpenAI's 39.7%. Anthropic's monthly adoption gain of 1.1 percentage points was nearly five times OpenAI's 0.23-point increase, though the period of rapid ascension appears to have slowed. Despite the overall lead, Anthropic's flagship Fable 5 model accounts for only 6% of Anthropic's tokens and 11.4% of model spend after roughly a month, with no rising share. The vast majority of Claude tokens run on Opus or Sonnet rather than Fable. Price is the cited constraint: Fable 5 costs approximately $10 per 1M tokens, about twice GPT-5.6 Sol. GPT-5.6 Sol represents 25% of OpenAI tokens and 23% of spend, while Fable 5 generated about 75% as much model-attributed spend as Sol in July. AI spending is heavily concentrated: the top 1% of businesses spent a median $7,400 per employee on AI in July, the top 10% spent $650, and the median firm spent $11.95 per employee, with the top 10% of businesses accounting for most total spending.
How it developed
WSJ reports Anthropic targeting September or early-October IPO ahead of OpenAI; annualized revenue at $47B driven by Claude Code.
Sources
Want this in your inbox?
I send a short email each morning with the stories that moved. If you would rather just read here, that works too.
Subscribe free