The token bill comes due: Inside the industry scramble to manage AI’s runaway costs - BERITAJA

Across the industry, companies are starting to balk astatine the value of AI. Uber blew through its full 2026 AI coding fund by April. Microsoft revoked its developers’ Claude Code licenses months aft enabling them. A Priceline worker told TechCrunch that a regular Cursor statement renewal came backmost 4-5x much expensive.

Even though per-token prices person fallen, the push for much AI take and progressively autonomous agents person driven token depletion higher and higher. Companies that gorged themselves successful early 2025 connected all-you-can-eat subscriptions are now scrambling to understand wherever their money is going, propulsion backmost spending, and fig retired whether they could salvage immoderate ROI from the wreckage of their budgets.

Meanwhile, a marketplace is forming to meet them there. Startups, established vendors, and a caller standards assemblage are each racing to springiness companies the devices and connection to way what they spend.

“Six months ago, I would person a speech pinch a customer and it would beryllium each about ‘What could it do? Is it bully enough?’” Alexander Embricos, OpenAI’s caput of enterprise, told TechCrunch astatine an arena successful New York City this week. “Our conversations are ne'er about that now. Now the conversations are about, ‘hey, we’re spending truthful much. What visibility do you have? What auditability do you have? What token controls do you have? What is the ratio of your models?’”

It’s against this backdrop that the Linux Foundation this week unveiled plans for the Tokenomics Foundation, a caller standards assemblage that intends to instill the aforesaid costs subject about AI tokens that FinOps did for unreality spend.

“In April and May, I started proceeding from companies: ‘Oh my god, we are 3x complete our full 2026 token fund and it’s only April,’” J.R. Storment, executive head of the FinOps Foundation, a task nether the Linux Foundation, told TechCrunch. “We started proceeding existential crises, and the full speech shifted from tokenmaxxing and ‘go fast’ to ‘we request guardrails, really do we power this?’”

The cries heard information the tech world followed fervent demands from CEOs pushing their teams to usage the champion models and move fast, costs beryllium damned. New models released successful November for illustration Anthropic’s Claude Opus 4.5, OpenAI’s GPT-5.1, and Google’s Gemini 3 Pro brought important improvements to agentic tools, which person multiplied consumption. It’s really 1 institution reportedly recovered itself pinch a $500 cardinal Claude measure aft forgetting to group usage limits for employees. 

“It’s for illustration the crack-cocaine epidemic,” says Chris Reed, elder head of IT finance astatine Priceline, erstwhile asked about the pricing rumor successful utilizing AI. “They fto you effort it to get you hooked connected it, and now you’re benignant of beholden to it.”

Vitaly Gordon, CEO of engineering operations level Faros AI, said he precocious said to a CTO who told him: “One of my engineers spent $40,000 connected tokens past month, and I genuinely don’t cognize whether I should extremity him aliases should I spell and show everyone other to beryllium for illustration him.“

A March survey by Faros recovered that among 20,000 developers, output was rising, but truthful were bugs and rewrites. Jellyfish, an engineering guidance platform, likewise recovered engineers who utilized the about tokens were about doubly arsenic productive than those who utilized AI less, but they spent 10x the number of tokens to get there.

Nicholas Arcolano, caput of investigation astatine Jellyfish, told TechCrunch via email that expenditure connected AI is exploding successful ample portion owed to agentic features, pinch per-developer depletion rising about 18.6x successful 9 months. All successful all, these stats make the productivity lawsuit murkier than the spending suggests.

“Whether utmost walk pays disconnected comes down to the eventual business worth of shipped codification (e.g. revenue), which about companies still can’t measure,” Arcolano said.

At slightest immoderate of that measurement rumor is the sheer standard astatine which AI is being utilized today.

“Tracking unreality costs is simply a hundreds-of-millions-of-rows-a-month information problem,” Storment said. “Tracking token costs is simply a trillions-of-rows-a-month information problem. You can’t conscionable instrumentality that into immoderate spreadsheet aliases moreover basal tool. You’ve sewage to fundamentally rethink your tooling, your specs and your accounting systems to do that.”

At Priceline, Reed is already seeing discrepancies. He noted issues betwixt a vendor’s reported usage and Priceline’s soul data.

“I started my profession successful telecom disbursal management, and I’m seeing each the aforesaid parallels, from telecom to unreality to AI,” he said. “Anytime you present thing new, it’s ripe for billing errors and audit and optimization opportunities.”

A marketplace is opening to shape about this problem. There are the pure-play companies, for illustration Pay-i, which tracks, measures and optimizes the costs and capacity of GenAI investments. Paid, meanwhile, lets developers way costs, measurement usage and measure users based connected existent worth alternatively than subscription fees.

Then location are companies for illustration Jellyfish, Waydev and Faros AI, which each supply AI supplier monitoring to beryllium the ROI of developer tools. Storment says about of the 180 vendors wrong the FinOps Foundation are leaning towards this space. 

Companies pinch existing distribution are besides adding caller features to capitalize connected this caller market. Ramp has precocious moved into AI walk management; Datadog and New Relic person tacked connected services for illustration unreality costs management, token-level observability, and GPU monitoring. At the FinOps X convention adjacent week, AWS is expected to present caller financial guidance features geared toward endeavor AI spending.

Tiffany Luck, a partner astatine NEA, thinks token ratio and observability will apt beryllium added successful astatine the “harness aliases app layer.” She pointed to Factory, a startup that makes AI agents for enterprises, which this week launched a exemplary router that automatically picks the correct exemplary for each task. 

Gordon expects frontier labs and different exemplary providers to adopt OpenRouter-style optimization to thrust queries to the cheapest models — a inclination already showing up connected endeavor Claude bills. 

“The financial study for really overmuch you walk connected Anthropic, moreover if you telephone the Opus model, immoderate of the walk will beryllium connected Sonnet aliases Haiku, because they are smart capable to do it,” Gordan said. “I deliberation this will go much and much of a thing.”

But each these devices are being built without a communal connection aliases shared definitions for really overmuch a token costs, what it produces, and really to comparison walk crossed vendors. That’s wherever the Tokenomics Foundation hopes to beryllium useful.

The Foundation is building a canonical meaning and model for “tokenomics;” unfastened standards, specifications and metrics for AI token usage and billing; arsenic good arsenic caller metrics for AI economics, for illustration cost-per-intelligence aliases tokens-per-watt. It besides plans to specify metrics crossed token mill effectiveness and depletion efficiency. The group is readying a general motorboat successful July, and is about to denote much members astatine the FinOps X convention adjacent week. 

“Token economics is fundamentally much absurd and opaque than thing we’ve managed astatine this standard before,” Nishant Gupta, main readiness serviceman astatine Salesforce, said successful a statement. “It requires a different operational musculus than the 1 the manufacture built for cloud.”

That said, Goldman Sachs projects world token usage to multiply by 24 times by 2030. The companies already complete fund request solutions now, and the foundation’s first deliverable is still months away.

“Maybe we created a steam engine, but we still haven’t figured retired the assembly line,” said Gordon.

According to Arcolano, the smart move is broad, mean adoption. 

“The champion ROI comes from moving the wide mediate from debased to mean usage, not pushing dense users higher,” he said.

Russell Brandom and Tim Fernholz contributed to this reporting.