The Linux Foundation plans to launch the Tokenomics Foundation, a new group focused on open standards for managing AI costs.
Created in partnership with the FinOps Foundation, it will develop standards, benchmarks and best practices for the economics of AI infrastructure.
The move comes as companies push generative and agentic AI projects beyond trials into wider deployment, prompting closer scrutiny of token-based pricing models. Tokens, used to measure consumption in many AI services, are taking up a growing share of technology budgets as businesses expand their use of large language models and other AI systems.
Jim Zemlin, Chief Executive Officer of the Linux Foundation, said the industry lacked a neutral body to create common measures for token spending across suppliers.
"As enterprises move generative and agentic AI workloads from pilot to production, tokens have become the new unit of technology spend. Measuring and benchmarking token efficiency across different models and vendors is critical to how organizations make business decisions, but until now, there was no neutral home to develop the standards needed to measure token economics transparently across the entire supply chain. The Tokenomics Foundation provides that neutral home, ensuring these standards remain open and community-driven," Zemlin said.
Backers of the planned initiative include Accenture, Booking.com, Flexera, Google Cloud, IBM, JPMorganChase, KPMG, Microsoft, Oracle, Salesforce, SAP and ServiceNow.
Cost pressure
The organisation is being formed as spending on AI infrastructure rises sharply and the declines in per-token pricing seen in recent years begin to ease. Prices for newer model tokens are rising, heightening concern among large buyers over how to compare suppliers and control spending.
Research cited by the organisation from Goldman Sachs projects global token usage will rise 24-fold between 2026 and 2030 to 120 quadrillion tokens a month. It also pointed to forecasts of more than USD $1 trillion in AI infrastructure investment through 2027, while the inference market is projected to grow from about USD $106 billion in 2025 to USD $255 billion by 2030.
J.R. Storment, Executive Director of the FinOps Foundation, said token pricing had moved beyond a narrow technical issue.
"Token costs and efficiency have become a CEO-level concern, not an engineering footnote. But naming the problem isn't solving it. The Tokenomics Foundation gives the industry a neutral home to define the standards, the specifications, and the discipline that will determine how much companies benefit from the inference era. In the same way FinOps created a shared discipline for cloud spend, Tokenomics will do it specifically for AI and related token costs," Storment said.
The planned structure includes a Governing Board to set direction and allocate funding, along with a Technical Committee to develop specifications, benchmarks and frameworks. The work will also support the expansion of the FOCUS specification into token-based spending models.
Buyer and supplier focus
The foundation is intended to cover both sides of the AI market: large enterprises seeking vendor-neutral ways to assess token consumption and cost, and model developers, cloud providers and newer infrastructure suppliers that need common methods for presenting pricing and efficiency.
Several early supporters framed the issue as one of financial control, not just technical performance.
"We work with thousands of enterprises reinventing themselves around AI, and the hardest conversation is no longer whether to adopt it but how to prove the return. Token spend is climbing fast and the discipline to govern it has not kept pace. When the bill arrives, companies face a tough choice between pouring more money in, or having to pull back and risk slowing innovation. Open, vendor-neutral standards for token economics give our clients a common language to manage that spend and quantify the investment and its returns. That is the gap the Tokenomics Foundation fills, and we are excited to help build it," said Mike Eisenstein, Managing Director, Accenture.
Booking.com linked the issue to operating AI systems at consumer scale.
"We serve travelers at enormous scale, and generative AI now touches everything from our Trip Planner to the agentic tools handling millions of customer and partner conversations. At that volume, the economics of every token matter, and small differences in efficiency compound into very large numbers. We need transparent, comparable ways to measure token cost and performance across models and providers so we can keep delivering value to travelers sustainably. The Tokenomics Foundation gives the industry the neutral standards to do that, and we are proud to support it," said Chris Reed, Senior Director, IT Finance, Booking.com.
Flexera pointed to its own cloud spending research as evidence that AI is changing cost management patterns.
"Our 2026 State of the Cloud research found cloud waste rising for the first time in five years, stemming in part from the surge in AI workloads. The real challenge for organizations is no longer just adoption, it's understanding spend and controlling AI costs to make it sustainable to run at scale. While teams are getting better at cloud financial management, token-based pricing behaves differently, and most still lack the benchmarks to know whether they are paying a fair price for the value they receive. A neutral, community-built standard is essential to give buyers a true picture, and that is why Flexera supports the Tokenomics Foundation," said Jay Litkey, Senior Vice President of Cloud and FinOps, Flexera.
IBM also argued that standard-setting should not be left to any one supplier.
"As tokens become the common currency of AI, enterprises need better ways to measure and manage their value, efficiency and cost across an increasingly complex ecosystem of models and platforms. As buyers choose among a growing range of models and deployment options, they need open, trusted standards to compare cost and efficiency across all of them. No single provider should define those benchmarks. That is why IBM supports the Tokenomics Foundation and the move toward a neutral, community-owned foundation for this work," said Bill Lobig, Vice President of IBM Apptio, IBM.
"Managing cost and performance is top of mind for every organization scaling AI today. ServiceNow lives this from two angles: governing and controlling AI at scale while managing costs internally, and building solutions that solve this problem for our customers. That dual perspective makes clear why the Tokenomics Foundation matters. The industry needs a central place to develop the standards and framework that turn AI spend into a strategic, accountable investment," said Dinesh Sonawane, Vice President, Technology Business Management and FinOps, ServiceNow.