![](\"https:\/\/fortune.com\/img-assets\/wp-content\/uploads\/2026\/09\/GettyImages-2263283192-e1788984401697.jpg?w=2048\")
<\/p>\n<\/p>\n
Almost everything in the AI economy is a bet about the future.<\/p>\n
When Nvidia reports its quarterly earnings, its backlog \u2014\u00a0orders planned but not yet filled \u2014 matters nearly as much as revenue. Anthropic and OpenAI\u2019s IPO chatter and \u201cvaluations\u201d are bets on what they\u2019ll earn as much as decades from now. Companies like Coreweave finance data centers before they have tenants. Everyone is pushing capital through the same bottleneck, roughly 2% of GDP a year, on a simple premise: demand for AI compute is close to infinite, so either businesses will pay more for smarter models, or they\u2019ll use so much of them that it won\u2019t even matter.<\/p>\n
Jensen Huang, CEO of Nvidia, calls that the \u201ctwo exponentials\u201d driving the price of AI compute; models are growing more complex, and more people and agents use them. Either way, the idea is that the labs capture that surplus and send it back through the ecosystem to cover their debts. There\u2019s just one problem: as impressive as the new model releases are, they don\u2019t seem to be causing sustained spikes in the price of AI compute\u2014in fact, the AI token is getting cheaper, fast.\u00a0<\/p>\n
That\u2019s according to new data from Ramp, the corporate spending platform, published Wednesday, showing the effective price that American businesses pay per a million tokens has fallen about 41% from its peak in March, from $1.15 to 68 cents. The share of usage going to frontier models is dropping, too; about 53% in early August to 45% by September. And the top 1% of spenders, the cohort that drives about 80% of OpenAI and Anthropic\u2019s enterprise revenue, cut per-employee spend by nearly 10% in August.\u00a0<\/p>\n
It\u2019s not a disaster or the bubble bursting but it is a \u201ccrack in the AI thesis,\u201d Ara Khazarian, the Ramp chief economist who runs the Index, told Fortune. Rather than unleashing a gush of demand for the best models, tokens are starting to be priced more like a commodity\u2013 as interchangeable as salt or wheat. And commodity owners aren\u2019t valued at $2 trillion. Morgan Stanley has flagged vulnerability for up to $300 billion in bonds financing neocloud buildouts\u2014CoreWeave-style companies that borrowed to build data centers before signing tenants\u2014if token prices don\u2019t keep up.\u00a0<\/p>\n
Ramp isn\u2019t the only one flagging the trend. Citadel Securities noted in June that a separate measure, Silicon Data\u2019s LLM Expenditure Index, started to fall because of a \u201cbifurcation\u201d between frontier AI, concentrated among the few tech-heavy firms that can afford it, and the \u201ceveryday\u201d AI the rest of the economy runs on.<\/p>\n
\u201cYou have multiple metrics now starting to move in a negative direction,\u201d Kharazian said.<\/p>\n
He said that the price decline reflects a mix of labs being forced to cut prices\u2014OpenAI slashed the cost of its GPT-5.6 Luna model by 80%, and Anthropic announced its own cuts last month\u2014and customers trading down to cheaper and simpler models. Which makes it threatening, he added, to anyone \u201cwho\u2019s expecting a full dream scenario where the AI companies grow with nothing curbing their enthusiasm.\u201d<\/p>\n
That was the mood in the Spring, as \u201ctokenmaxxing\u201d entered the tech lexicon, the media told stories of token-usage dashboards and Nvidia\u2019s Huang insisted a $500,000 engineer should burn $250,000 a year in tokens. But by the summer, cost discipline set in; Amazon and Meta killed its own leaderboards in May, while Microsoft cancelled Claude Code subscriptions. Khazarian said he\u2019s now hearing the opposite of tokenmaxxing from businesses: companies are imposing defaults that steer employees away from frontier models entirely.<\/p>\n
\u00a0\u201cCompanies are increasingly starting to use Terra and Sonnet,\u201d he said; the mid-tier models that are \u201chighly performant and also cheaper.\u201d Ramp\u2019s top 1% of firms, the most AI-intensive in the country, now spend about $7,200 per employee per month on AI\u2014roughly a third of Huang\u2019s target, and tapering off.<\/p>\n
Some analysts blame the rise of open-source models, a hot-topic issue a few weeks ago for AI companies that has now faded to the background among the high tenor of AI discourse drama. Only 3.6% of businesses on Ramp\u2019s platform use open source or Chinese models. And even if they did, there\u2019s good evidence that Deepseek, Tencent, Alibaba, and other major Chinese companies are fighting a brutal pricing war that\u2019s depressing token prices too.<\/p>\n
So it\u2019s an international phenomenon. Back home, OpenAI and Anthropic are in a lopsided competition, Kharazian said. Since Aug. 1, OpenAI\u2019s effective price has fallen 38%, to 48 cents; Anthropic\u2019s has fallen 22%, to 90 cents. Anthropic has charged nearly double of what OpenAI has all year and has held down a floor near 90 cents since June, but OpenAI\u2019s price has kept falling. That suggests Anthropic \u201cprobably has some pricing power,\u201d Kharazian said, but \u201cthat edge is wearing down\u201d as OpenAI takes a growing share of tokens on price.<\/p>\n
The token pricing patterns tell a similar story; Anthropic\u2019s price spiked when Fable 5 launched in March, and OpenAI spiked with Sol in July. But they weren\u2019t consistent, and prices eventually faded back down. \u201cIt\u2019s not that businesses aren\u2019t willing to pay high prices,\u201d Kharazian said. \u201cPrices are relative to the other products available on the market,\u201d and the mid-tier products \u201care also really good.\u201d<\/p>\n
OpenAI\u2019s finance chief described that dynamic on Tuesday as she spoke at a Goldman Sachs conference. CFO Sarah Friar said the company had cut the price of its GPT-5.6 Luna model by 80% since its launch, cheaper than some Chinese open-source weight models. She added that she\u2019d like to get away from token counting altogether, and move enterprise customers to paying only for completed work. They claim that OpenAI\u2019s enterprise revenue grew 32% from June to July, she said, but that growth seems to come from increased share.<\/p>\n
\u201cI would love,\u201d Friar said, \u201cto get us away from token-counting.\u201d<\/p>\n
#latest #crack #thesis #trilliondollar #boom #Tokens #cheaper<\/p>\n","protected":false},"excerpt":{"rendered":"
Almost everything in the AI economy is a bet about the future. When Nvidia reports…<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[245],"tags":[353,1715,2657,7805,1552,4272,1837,335,406,21058,7978,12847],"_links":{"self":[{"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/posts\/15689"}],"collection":[{"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=15689"}],"version-history":[{"count":0,"href":"https:\/\/stock999.top\/index.php?rest_route=\/wp\/v2\/posts\/15689\/revisions"}],"wp:attachment":[{"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=15689"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=15689"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/stock999.top\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=15689"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}