US Corporate AI Spending Reaches New Peak in October as Token Prices Drop 21 Percent
Ramp data shows US corporate AI spending rose 1.65% in October 2026. While token prices fell 21%, top spenders paid $10,200 per employee monthly while typical companies spent just $16.32.
Corporate spending on artificial intelligence in the United States reached a new peak in October 2026, though overall month-over-month growth slowed to 1.65 percent, according to recent payment data from financial services provider Ramp. While aggregate expenditure continues to climb, investment remains heavily concentrated among a small tier of high-volume business users, even as average token costs fall sharply.
Concentration of Spend and Vendor Adoption
The Ramp AI Index, which analyzes transaction metrics from more than 85,000 U.S. businesses, highlights a significant gap between heavy AI users and average corporate buyers. The top one percent of companies by AI spending paid a median of $10,200 per employee per month in October, representing a nine percent increase. By contrast, a typical company spent just $16.32 per employee, reflecting a modest increase of less than one dollar from the previous month. Ramp noted that its customer base leans toward tech-savvy, growth-oriented, and larger businesses, which likely overstates broader adoption across the general economy.
Vendor performance varied across the platform during the month. Business spending on OpenAI grew 2.1 percent month over month, outstripping Anthropic's 0.5 percent spending growth. However, Anthropic achieved a higher overall adoption rate of 44.8 percent among tracked companies (up 1.1 percentage points), compared to OpenAI's steady adoption rate of 40.6 percent. SpacexSI posted the fastest software spending growth on Ramp at 2.2 percent, holding a 6.3 percent market share driven primarily by its coding tool, Cursor. Across all tracked vendors, corporate adoption of AI services rose to 56.8 percent.
Falling Token Prices and Revenue Risks
Despite high overall adoption rates, financial revenue for major model creators remains highly concentrated. According to Ara Kharazian, chief economist at Ramp, one percent of enterprise clients drive 80 percent of total revenue for OpenAI and Anthropic. This high-spending cohort consists mostly of venture-backed firms in technology, finance, and technical services, while expansion into other business sectors remains limited.
Concurrently, processing costs are declining. The average price per million processed text tokens dropped 21 percent in October. Usage volume is expanding faster than total monetary spending as competitive pressures build between OpenAI, Anthropic, and open-weight models from China.
Departmental Spending Shifts Toward Support
Departmental allocation of AI budgets is also undergoing a transition. Software development teams accounted for 72 percent of all tracked AI spend in October, though their overall expenditure has dropped from highs recorded over the summer. Meanwhile, customer support and customer success departments recorded substantial gains, increasing their share of corporate AI spend from 0.8 percent in June to 13.3 percent in October, likely driven by the deployment of customer service chatbots. Executive leadership teams also posted small spending increases.
Kharazian pointed out that customer support use cases generally favor lower-cost, faster models over flagship models. As new enterprise customers onboard for customer service applications, providers may experience lower profit margins, even though expensive flagship models—such as Opus, Fable, Sol, and Astra—still represented 69.5 percent of total spending in October.
What it means for developers
For software engineers and technical leads, falling token prices mean that building and running AI-powered applications is becoming significantly more cost-effective. Even though development teams still represent 72 percent of total corporate AI spend, declining token costs allow engineers to process higher transaction volumes without a proportional increase in infrastructure spend.
At the same time, managing diverse AI workflows—ranging from coding assistants like Cursor to automated support bots—often requires interacting with different providers. Developers can try top AI models cheaply through one API at https://apixoai.online to streamline multi-model integration without maintaining separate vendor billing systems.
As corporate spending expands into customer support and operations, developers will likely be tasked with building lightweight, high-throughput pipelines that leverage lower-cost models alongside high-end flagship systems.
Source: US corporate AI spending hits a new high in October, but growth slows to 1.65 percent — The Decoder. Written by the Apixo team from that report.
One key for Claude, GPT, GLM, DeepSeek and more. Pay per token with crypto.
Get your API key
