THE CRUNCH
Companies are using more AI than ever, but they are not paying more for it. Ramp's data shows token spend grew just 1.5% across August and September, down sharply from 42% month-on-month growth in July, even as usage volumes kept rising. The timing is awkward for Anthropic, which filed its S-1 prospectus in late October to list on Nasdaq under the ticker ANT after reporting an $80.7 billion net loss on $16.3 billion of revenue in the first half of the year, because the figures suggest AI revenue may not keep growing at breakneck speed.
Ramp's lead economist Ara Kharazian attributes the slowdown not to fading demand but to a price war. OpenAI halved API prices for GPT-6 Sol and GPT-6 Luna in September, and Anthropic answered with significant cuts to Fable 5.1 and Opus 5.5, a decision it said in a recent filing slashed revenue growth by about $600 million. Enterprise discounts meant to lock in spend are also hitting revenue targets, making it what Kharazian calls a buyer's market.
The heaviest users are wobbling too. Kharazian said the top 1% of firms, the most AI-intensive spenders, dipped just below $10,000 per employee per month from their summer peaks, and that this cohort alone is sufficient to slow revenue entirely. Median firms spend only $12 to $15 per employee monthly and the top 10% $600 to $1,000, and he is not convinced typical businesses such as doctor's offices or car dealerships will ever reach the top cohort's levels.
Competition compounds the pressure. Meta and Nvidia are introducing enterprise offerings with cheap or open source models, and a lightweight tool called Jev, which routes simple tasks to a few predetermined outputs, is increasingly among the top performers on Ramp. Companies are also getting thriftier, using defaults that avoid expensive models on long-running tasks and offloading work to cheaper light models via routers. October brought a partial rebound, with token spend up 8.9% from September 23 to October 23, but Kharazian said headwinds remain strong.
Notion co-founder Akshay Kothari posted internal data in October that matched Ramp's picture, summarising it as "tokens go brrr. costs go ahhh."