The AI Spending Plateau
Companies are using more AI than ever, but they are not spending more on it.
A new report from Ramp shows AI spending flatlined over late summer even as token volumes rose, a sign that price competition and savvier customers may be capping revenue for OpenAI and Anthropic. According to Ramp’s data, token spend only grew 1.5% across August and September, compared to a 42% growth rate in July.
As Anthropic gears up for its IPO, the data suggests investors might be wrong to assume AI companies will keep growing at breakneck speed.
Market Dynamics and Efficiency
Almost the entire spend on AI is from venture-backed, high-growth tech companies. If that spending slows down due to market factors, it could lead to cascading effects across the industry.
The top 1% of firms, identified as the most intensive AI users, have seen their per-employee spending dip below summer peaks of $10,000. This slowdown among high-volume users is sufficient to impact overall revenue growth.
Ramp data shows growth in spending among median and top 10% firms, but it is not enough to offset the plateau. Median firms spend $12-$15 per employee per month, while the top 10% spend $600-$1,000.
While demand for AI coding agents remains strong in software and finance, it is unclear if other industries like healthcare or retail will reach the same spending levels.
The era of hypergrowth may be over as the market realizes there is limited room to squeeze more revenue from the top tier of users.
October data showed a slight re-acceleration with token spend up 8.9%, but analysts suggest the headwinds on long-term spend growth remain strong.
The Impact of Price Wars
The slowdown is not a demand problem, as AI volumes continue to rise. Instead, it is driven by the competitive nature of the marketplace.
OpenAI and Anthropic have both engaged in aggressive price cuts for their API models. Anthropic noted in a recent filing that these reductions slashed revenue growth by approximately $600 million.
This dynamic has shifted the industry into a buyer’s market.
Unlike other tech sectors where subsidies lead to eventual price hikes, the AI market faces intense competition that prevents similar consolidation.
Meta, Nvidia, and various open-source providers are introducing cheaper models, further pressuring the pricing power of leading AI firms.
For many use cases, companies are shifting to lightweight tools that provide simple outputs, making already inexpensive models even more cost-effective.
Operational Thriftiness
The ability to complete workloads at a lower cost is driving the spending slowdown, as companies become more proficient at managing their AI usage.
Advanced users are implementing better controls, such as using default settings to avoid expensive models for simple tasks and using routers to offload work to cheaper alternatives.
Internal data from companies like Notion confirms this trend, with token usage rising while overall costs remain contained.
The assumption that AI usage will always lead to linear revenue growth for model providers is breaking down as customers find ways to do more with less.
Investor Sentiment and Future Outlook
Reports indicate that potential valuations for companies like Anthropic and OpenAI have been adjusted downward, reflecting investor caution regarding revenue trajectories.
Ramp’s dataset is likely informing these investor concerns, though some hope remains for new, high-value use cases to drive future growth.
The current market relies on the hope that a new technological revolution, similar to the emergence of AI coding agents, will emerge to justify continued spending increases.
While personal agents are being positioned as the next major product, there is skepticism regarding whether they will be as economically productive as coding agents.
Based on current commercially available products, the trend of slowing AI spend appears to be a significant factor for the industry to navigate.
