New data from Ramp indicates that corporate spending on AI leveled off in late summer even as usage continued to rise. Ramp reports token spend increased only 1.5% across August and September, after a 42% month-over-month rise in July. October showed a partial recovery with token spend up 8.9% from September 23 to October 23, according to Ramp.
Spending patterns and company cohorts
Ramp’s analysis finds that most AI expenditure comes from venture-backed, high-growth tech firms, and that the top 1% of users — the most AI-intensive companies — have eased back from their summer peaks. Ramp’s lead economist Ara Kharazian said those elite firms fell slightly below the roughly $10,000 per-employee level they reached at peak usage, and that a slowdown among that cohort can meaningfully affect revenue trends even as volumes rise.
Median firms, Ramp shows, spend about $12–$15 per employee per month, while the top 10% spend $600–$1,000. Kharazian cautioned that many industries may never reach the extreme per-employee spending seen in software and finance.
Price cuts, competition and efficiency
Kharazian attributes the spending slowdown not to lack of demand but to price pressures and greater cost-efficiency. OpenAI cut API prices in half in September for GPT-6 Sol and GPT-6 Luna, and Anthropic reduced prices for Fable 5.1 and Opus 5.5; Anthropic said in a filing those reductions lowered revenue growth by roughly $600 million. Both firms have also offered enterprise discounts intended to secure longer-term contracts.
Ramp’s economist noted broader market forces: new low-cost enterprise offerings from Meta and Nvidia, open models from multiple vendors, and lightweight tools such as Jev, which can route simple tasks to cheaper models. Companies are also optimizing usage by enforcing budgets, using cheaper default models, and routing workloads to less expensive options.
Notion co-founder Akshay Kothari shared internal data that aligns with Ramp’s findings, summarizing the trend as “tokens go brr. Costs go ahhh.” Bloomberg reporting has also tied these developments to lower valuations for Anthropic and OpenAI and reduced fundraising targets. Ramp’s dataset and Kharazian’s analysis suggest that while AI use can increase, revenue growth faces notable headwinds from pricing and competition.
Original source: TechCrunch AI