Key facts
- Ramp's lead economist Ara Kharazian said the top 1% of firms spent just below $10,000 per employee in August and September, under summer peaks.
- Median firms on Ramp spend $12 to $15 per employee per month; the top 10% spend $600 to $1,000, per Ramp.
- Anthropic said in a filing that its price cuts reduced revenue growth by about $600 million.
- Kharazian said OpenAI and Anthropic price cuts and enterprise discounts have made AI cheaper to use, creating a buyer's market.
- Kharazian said the slowdown is not a demand problem because AI token volumes are still rising.
AI token spending at companies tracked by spend-management firm Ramp grew 1.5% across August and September 2026, after growing 42% in July from the previous month, Ramp's lead economist Ara Kharazian told TechCrunch in a report published November 6.
Kharazian said the top 1% of firms, the heaviest AI users, spent just below the $10,000 per employee they had reached at their summer peaks. Median firms spend $12 to $15 per employee per month and the top 10% spend $600 to $1,000, according to Ramp data. He said a slowdown among that top cohort is enough to slow revenue overall even as token volumes rise.
Kharazian attributed the flattening to price competition rather than weak demand. OpenAI cut API prices in half for GPT-6 Sol and GPT-6 Luna in September, and Anthropic cut prices for Fable 5.1 and Opus 5.5, a decision the company said in a filing reduced revenue growth by about $600 million. Both companies have also offered enterprise discounts, which Kharazian said are hitting revenue targets.
He also pointed to cheaper enterprise and open-source models from Meta and Nvidia, open model companies in China and the United States, and Jev, a tool for choosing between predetermined outputs that he said is increasingly among Ramp's top performers. Customers are also imposing budgets and routing work to cheaper models, he said.
Ramp data showed an 8.9% rise in token spend from September 23 to October 23, which Kharazian said did not remove the pressure on spending growth. Notion co-founder Akshay Kothari posted internal data that Kharazian said matched Ramp's findings.
Kharazian said almost all AI spending comes from venture-backed, high-growth technology companies, so a slowdown would hit them at the same time. He was skeptical that personal agents, which he called less economically productive than coding, would restore the growth rate.