Incident sheet · /enterprise-ai
- System
- Enterprise AI
- Filed
- Read
- 4 min · 813 words
- Severity
- Routine
Tokenomics Foundation Wants to Make AI Bills Explainable
Launched August 2026 under the Linux Foundation, the Tokenomics Foundation publishes open standards, a 7-layer cost stack and Big-T notation to tie AI spend to business outcomes.
Event log
- The Tokenomics Foundation launched in August 2026 as a Linux Foundation project, working alongside the FinOps Foundation.
- Its cost-data standards extend FOCUS, the FinOps billing format, so AI providers report usage uniformly.
- One token roughly equals 4 characters, or 0.75 English words.
- A seven-layer stack maps AI costs from energy and GPUs up to user demand.
- Big-T notation ranks token-cost growth from T(1) cached answers to T(∞) unbounded loops.
The Tokenomics Foundation, a Linux Foundation project launched in August 2026, has set out to close a gap most enterprises know firsthand: teams can quote their monthly AI bill, but very few can say which use actually drove it.
The organization's answer is a set of open, vendor-neutral standards — definitions, frameworks, cost-data formats and benchmarks — designed to move AI from what it calls "experimental spending" to "accountable operations."
In conversations across the industry, one question keeps repeating: "We're spending on AI, but what are we getting for it?" Companies now spend real money on tokens, API calls and GPUs, yet few can connect that number to a business result. The Foundation's tagline captures the goal: don't just count tokens, get value from the total cost of AI.
What is AI tokenomics?
Every prompt and every response is split into tokens, and each one costs something. As a rough rule, one token equals about 4 characters, or 0.75 words, in English.
But the price per token is rarely the problem. The problem is how many tokens a system quietly consumes. AI tokenomics is the practice of managing that economy: understanding the cost structure and connecting it to business outcomes, from the power and hardware that produce tokens to the value they create.
One clarification the Foundation stresses: this has nothing to do with crypto. In crypto, "tokenomics" means the supply and incentives of a coin. Here, it means the cost and value of running AI.
The takeaway: unit price matters less than volume, and volume is usually unmanaged.
Who is behind it and what do they ship?
The Tokenomics Foundation operates as a Linux Foundation project and works alongside the FinOps Foundation, which previously did the same job for cloud spending. Its members include large AI buyers — banks and retailers — plus cloud and model providers.
It sells no product and ships no cost-tracking tool. It functions as a standards body whose working groups publish:
- Definitions, so everyone means the same thing by input, output, reasoning and cached tokens
- Frameworks, including the two described below
- Standards for cost data, mainly extending FOCUS, the FinOps billing format, so AI providers report usage in the same way
- Benchmarks, training and a certification for practitioners
The actual measuring stays with market tools such as AI gateways and observability platforms; the Foundation's job is to make sure they all speak the same language.
In short: shared vocabulary first, tooling stays commercial.
Where do AI costs actually come from?
The Foundation's first framework is a seven-layer stack mapping everything that shapes the cost of an AI answer, from electricity at the bottom to the user's request at the top:
- 6. Demand — who is asking, for what, and how much context they send
- 5. Access and control — the gateway in front of the models: routing, budgets, guardrails
- 4. Model selection — which model runs, and whether it's the right size for the job
- 3. Inference — the software serving the model, including caching and agent loops
- 2. Infrastructure management — how much of the hardware is doing useful work
- 1. Hardware — the GPUs and data centers that turn power into tokens
- 0. Energy — the price and supply of electricity and cooling
Its value is simple: it shows you where to look before you start cutting. The big savings usually come from the middle and the top — using a smaller model for simple tasks, caching repeated work and sending less context.
How fast do costs grow? Enter Big-T
The second framework, Big-T notation, borrows from Big-O in programming. It describes how fast token use grows as more people use an application, ranked from cheapest to most dangerous:
- T(1), constant — the answer comes from a cache; no model is called at all
- T(n), linear — one model call per request; the healthy default
- T(n·k), multiplied — each request quietly triggers several model calls
- T(n·k·a), agent-multiplied — agents spawn other agents, and calls pile up fast
- T(∞), unbounded — a loop with no exit, which runs until someone pulls the plug
The multiplier is usually invisible. A user asks one question, and the app makes five or ten calls behind the scenes. Big-T gives teams a shared way to spot that before the bill does.
Why it matters now
Token prices keep falling, but AI bills keep rising — cheaper tokens lead teams to build bigger systems, with longer conversations, more documents per question and agents calling agents.
The practical starting point for any team: ask how many model calls one user request really triggers. That single number predicts the future bill better than any price list.
Most organizations today cannot connect AI spend to an outcome their CFO would accept, and that is precisely the gap the Foundation's standards, benchmarks and certification aim to close.
via infoworld.com/feed/ (Source)
More from Amara Osei
Show full bio
Correspondent covering regulation, policy and their impact on enterprise information technology and the leaders who run at CIO Monitor.
188 articles
The newsroom digest
One email a day with the stories that matter. No noise.
Cross-references
- Enterprise AI Bills Set to Outpace Falling Model Costs
- Gartner Sees AI Spending Up 54.2% in 2027 as Token Prices Fall
- MIT Report: 95% of Enterprise Generative AI Pilots Are Failing
- Goldman Sachs: Enterprise IT Spending Sentiment Weakens as AI Reshapes Budgets
- The Tokenizer Tax: AI Costs Up to 17x Higher in Non-English Languages