AI Spending Flatlines as Anthropic Nears Its IPO

Token spend on AI grew just 1.5% across August and September, according to spending data from corporate card platform Ramp — down from a 42% month-over-month jump in July. That stall lands at an awkward moment: Anthropic is gearing up for its IPO, and investors may be wrong to assume AI companies will keep growing at breakneck speed.

The core signal is a divergence. Companies are using more AI than ever, but they're not spending more on it. Ramp's lead economist Ara Kharazian attributes the gap to price competition and increasingly thrifty customers — dynamics that cap revenue for OpenAI and Anthropic even as usage climbs.

Who's actually spending?

The customer base is dangerously narrow. "Almost the entire spend on AI is from venture-backed, high-growth tech companies," Kharazian told TechCrunch. "That's fine as long as AI spend continues to rise, but if it starts to slow down because of something in the market or some other factor, they will all start to slow down simultaneously, and that could lead to other cascading effects."

The top 1% of firms — the most "AI-pilled" — spent just under $10,000 per employee at their summer peaks, and they stuttered slightly in August and September. Median firms spend $12-$15 per employee per month; the top 10% spend $600-$1,000. Growth among those groups isn't enough to fill the hole. A slowdown among the 1% alone is "sufficient to slow down revenue entirely, despite the fact that we're seeing rising volumes," Kharazian said.

"The era of hypergrowth may be over, and now that we know there may not be as much juice to squeeze out of the top 1%," he added. October offered a partial reprieve — token spend rose 8.9% from September 23 to October 23 — but Kharazian said the headwinds on spend growth remain strong.

What does this change for the model companies?

The answer is pricing power, and it's evaporating. In September, OpenAI cut API prices in half for GPT-6 Sol and GPT-6 Luna. Anthropic answered with significant reductions on Fable 5.1 and Opus 5.5 — a decision the company said in a recent filing slashed revenue growth by about $600 million. Both companies have offered enterprise discounts meant to lock in spend, but in reality these are hitting their revenue targets.

Kharazian frames it bluntly: it's now a buyer's market. The classic Big Tech playbook — subsidize expensive technology, corner the market, then raise prices like Uber and Lyft did — doesn't transfer. "Meta and Nvidia are also introducing enterprise offerings with pretty cheap models or open source models," he said, alongside open model companies from China and the United States. A new tool called Jev, a lightweight chooser between predetermined outputs, is now among the top performers on Ramp. "For a lot of AI use cases, you just need a simple Yes/No," Kharazian said. "So you take your already cheap models, you shift them over to Jev, and now your already cheap stuff gets even cheaper."

Customers are also getting better at controlling costs — using defaults that block expensive models for long-running tasks, imposing employee budgets, and routing workloads from standard models to cheaper light models via routers. Notion co-founder Akshay Kothari posted internal data matching Ramp's findings: "tokens go brr. Costs go ahhh."

Valuations already reflect it

Bloomberg reported Friday that Anthropic's potential valuation sank from $2 trillion, estimated in early October, to $1.2-$1.3 trillion, with a high "junk" credit rating expected. OpenAI cut its proposed valuation to $1.25 trillion from $1.4 trillion and slashed fundraising goals by about $10 billion. Kharazian says Ramp's dataset likely informs many of those investor concerns.

The remaining bull case rests on a new product wave. AI coding emerged only a year and a half ago and turned Anthropic and OpenAI into trillion-dollar companies; OpenAI, SpaceXAI, Meta, and startups like Instinct have all released personal agents. Kharazian is skeptical those agents — "not as economically productive as coding" — can deliver nonstop growth. "AI spend has started to slow down," he said, and until a commercially proven product arrives, model companies' revenue trajectories face questions the IPO market will have to price in.