Ramp: Top AI Firms Spend $7,500 Per Employee Monthly
Key insights
- Ramp's AI Index finds the top 1% of firms spend roughly $7,500 per employee monthly, growing 14.1% month-over-month.
- The median company spends just $11.38 per employee monthly, compared to $611 among the top 10% of spenders.
- An Nvidia executive and Mercor's CEO both report that AI spending at their companies already exceeds employee payroll costs.
Why this matters
The $7,500-per-employee-monthly figure from Ramp's AI Index gives boards and CFOs a concrete benchmark for what full AI commitment actually costs, a reference point missing from most budget conversations until now. The 14.1% month-over-month growth rate among the top 1% means the gap between AI-committed and AI-dabbling firms is compounding fast, with direct implications for product velocity and competitive positioning. The cases of Nvidia and Mercor, where AI spending already exceeds human payroll, signal that the central question for technical leaders is no longer whether to invest heavily but how soon compute costs will overtake headcount on their own balance sheets.
Summary
Ramp's AI Index shows a stark divide in corporate AI spending. The top 1% of spenders, labeled 'AI-pilled,' now spend roughly $7,500 per employee monthly on AI, growing 14.1% month-over-month.
That figure still sits below the roughly $16,000 monthly cost of a software engineer, though some firms have already crossed that line. An Nvidia executive said compute costs now exceed employee salaries at his company; Mercor's CEO reported internal AI agent token spending already surpasses their payroll.
Essentially: (Ramp, Nvidia, Mercor) are mapping what AI-first spending actually looks like in practice.
- The top 10% of companies spend around $611 per employee monthly; the median firm spends just $11.38.
- Top spenders blend multiple frontier models with open-source alternatives to manage costs.
- At 14.1% monthly growth, the gap between the top 1% and the median is actively widening.
The $11.38 median shows most companies are still watching from the sidelines.
Potential risks and opportunities
Risks
- Companies at the $11.38 median monthly AI spend face compounding productivity gaps versus AI-pilled competitors, particularly in software development and content-heavy workflows
- Firms scaling rapidly toward the $7,500-per-employee figure without measurable ROI metrics risk board-level spending backlash in the next one to two quarters as AI costs approach payroll line items
- Companies like Mercor, where AI agent token spending already exceeds payroll, face concentration risk if a frontier model provider raises prices or restricts API access
Opportunities
- Open-source model providers such as Meta and Mistral benefit directly as top spenders adopt a 'mix and match' strategy combining frontier and cheaper open-source alternatives to manage costs
- AI spend management and optimization platforms gain a clear sales pitch as per-employee AI costs at leading firms approach the roughly $16,000 monthly software engineer salary level
- Talent-light startups with high AI intensity can use the Ramp data to justify investor scrutiny of traditional headcount growth in favor of AI infrastructure spend
What we don't know yet
- Which specific AI vendors and frontier models are capturing the majority of the $7,500 per employee monthly spend among the top 1% of firms
- Whether the 14.1% month-over-month growth rate among top spenders reflects a sustained multi-month trend or a recent acceleration
- How Ramp defines the 'AI-pilled' cohort beyond spending thresholds, and whether its membership has remained stable or turned over as costs rise
Originally reported by techcrunch.com
Read the original article →Original headline: Ramp AI Index: Top 1% of Firms Spend $7,500 Per Employee Monthly on AI, Approaching Parity With Engineer Salaries