Live model
The token economy, audited.
Live model — estimated global AI activity
Not a measurementLive model
Estimated global activity, smeared from public prints. Tickers are a live model — not a utility meter.
AI tokens today
—
About 4.4 billion tokens every second — War and Peace, 7,800 times, every second.
Modeled estimate · smear clock. Method
Tokens per day, trillions
2026 · modeled series
Modeled, since you opened this page
0 tokens
Spend0
Queries0
kWh0
The five
Alphabet leads the year in this five. Nvidia still sets the market cap. Meta leads the month.
9 October 2026. Five US listings that set the AI trade: the chip toll and the four hyperscalers. Delayed session print, not a live quote and not a recommendation.
Nvidia · Accelerators · $5.6T. Week -0.2%, month +3.0%, year +19.7%.
8 October close about $230.48, down 2.9% on the day, roughly flat on the week, up about 3% on the month and about 20% on the year. Market cap about $5.6 trillion, still the largest of the five. The session followed a report that OpenAI’s September run rate was near $50 billion, not the higher figure circulating last month. A delayed print, not a trading signal. The $1 billion Genesis credit pledge is not in this price as cash received.
Alphabet · Search, cloud, Gemini · $4.3T. Week +3.0%, month +5.3%, year +44.2%.
About $348, down less than 1% on the day, up about 3% on the week. The year is still the outlier, about 44%, well ahead of Nvidia in this set. Class A print only. Google’s $150 million Genesis credit is a pledge, not a revision of the listing.
Microsoft · Azure, Copilot · $3.8T. Week +1.9%, month +6.3%, year 0.0%.
About $523, down 1.3% on the day. The month is about 6% higher. The year is about flat, the quietest of the five with Meta. The same day Labor said it would not process Microsoft’s permanent labor certifications, and Nadella received a technology medal. Neither is a closed case.
Amazon · AWS · $2.8T. Week +2.3%, month +0.7%, year +11.6%.
About $254, down 2.3% on the day, up about 2% on the week and about 12% on the year. AWS is the AI line that matters. The reported OpenAI revenue gap turns partly on whether partner cloud sales are counted. That is a reporting difference, not a closed Amazon deal.
Meta · Ads, open weights · $1.8T. Week -0.7%, month +10.3%, year -1.7%.
About $721, flat on the day, down less than 1% on the week, up about 10% on the month — the best month of the five — and slightly down on the year. The ad engine still pays for the open-weight models. No new company print overnight moves this row.
The desk
9 October 2026, morning desk
White House
Eleven firms pledged $2.4 billion in compute credits. Not an appropriation.
8 October White House fact sheet from the Science: A New Golden Age summit. The posted industry total is $2.4 billion in tools and compute credits for the Genesis Mission Consortium, for more than 15 federal agencies. Named shares: Nvidia $1 billion, AMD $500 million, OpenAI $200 million, Anthropic and Google $150 million each, AMP and Emerald AI $100 million each, AWS, Armada, Crusoe, and Micron $50 million each. Nvidia’s own newsroom says its $1 billion is spread over five years across university research, quantum, and cloud for government. Anthropic’s post says its $150 million is over three years, for Claude, Claude Code, and API credits. A pledge list, not a statute and not cash already spent.
OpenAI
September run rate near $50 billion, not the $70 billion figure.
8 October, Reuters from a person familiar, after the Financial Times. OpenAI told investors annualized revenue for September was almost $50 billion. The company had earlier signaled a figure approaching $70 billion. The person said the gap is mostly partner gross — sales through AWS and Google Cloud — which Anthropic counts and OpenAI does not. CNBC confirmed about $50 billion and said the investor update also cited 77% total run-rate growth in the quarter and 107% for enterprise. OpenAI did not comment to Reuters. A private investor print, not a filing.
Labor
Labor halted Microsoft’s green-card filings. A suspension, not a charge.
8 October, Reuters. Labor Secretary Keith Sonderling said the department will not accept new or process pending permanent labor certification applications involving Microsoft, Adobe, Cognizant, Infosys, Tata Consultancy Services, Wipro, HCL, and Capgemini. Vice President JD Vance tied the hold to alleged replacement of American workers. Microsoft is the largest PERM filer on Labor Department data. Nadella still received a National Medal of Technology the same day. An administrative suspension and a speech. Not a complaint and not a finding.
White House
Huang, Musk, Brin, and Su got the science medal. Nadella and Dell got technology.
8 October White House release from the same summit. National Medal of Science: Jensen Huang, Elon Musk, Sergey Brin, Lisa Su. National Medal of Technology and Innovation: Satya Nadella and Michael Dell. A ceremony. It does not change export rules, the 29 September accord, or Executive Order 14434.
On the tape
9 October 2026
Capitol
Eleven firms pledged $2.4 billion in Genesis compute credits
8 October White House fact sheet from the Science: A New Golden Age summit. The posted total is $2.4 billion in tools and compute credits for the Genesis Mission Consortium, supporting more than 15 federal agencies. Named shares: Nvidia $1 billion, AMD $500 million, OpenAI $200 million, Anthropic and Google $150 million each, AMP and Emerald AI $100 million each, AWS, Armada, Crusoe, and Micron $50 million each. Nvidia’s newsroom says its share is over five years. Anthropic says its $150 million is over three years. A pledge list. Not an appropriation, not a statute, and not cash already spent. The fact sheet also restates NSF and Energy putting more than $100 million into instrumentation and autonomous labs.
Capitol
Labor halted Microsoft’s permanent labor filings
8 October, Reuters. Labor Secretary Keith Sonderling said the department will not accept new or process pending PERM applications involving Microsoft, Adobe, Cognizant, Infosys, Tata Consultancy Services, Wipro, HCL, and Capgemini. Vice President JD Vance said the hold targets replacement of American workers. He did not publish a case file. A suspension announced at a press event, the same day Nadella received a medal. Not a complaint and not a court finding.
East
Beijing calls distillation ‘neutral’ after Anthropic’s clone claim
China’s commerce ministry rejected US distillation allegations and warned of retaliation. The same week Anthropic accused Moonshot and DeepSeek of diverting live queries onto Claude.
Compute
Global AI compute, expressed in a chip that is already a generation behind. The fleet is the story. The model is the exhaust.
Energy
29.6 gigawatts of AI data-center capacity is no longer a rounding error on a grid. It is a constituency.
Labor
Workers who can drive the tools are being repriced faster than schools can mint them. The agents that would replace them are still, mostly, a physics problem.
Census
AI accelerators · modeled estimate · stock
—
~2 / hour
Global operational AI compute, normalized to NVIDIA H100s. Nvidia still holds the majority; Google TPUs and Amazon Trainium make up much of the rest.
Growing by about one H100-equivalent every 1.5 seconds.
Source: Stanford AI Index 2026 — 17.1M H100-equivalents, 3.3×/year since 2022.
Electricity · modeled estimate · today
—
~5,920 / second
AI data-center IT capacity around 29.6 GW. We meter a 72% average load — training clusters spike, inference idles, cooling is extra.
About 21 GW continuous — a mid-sized country's evening demand.
Source: Stanford AI Index 2026 (29.6 GW AI data-center power capacity).
Agent-equivalents · modeled estimate · stock
—
~9 / hour
Not chatbots — the theoretical number of full-time knowledge workers the current token factory could replace if every token went to autonomous work. Tiny today. Not tiny in 2028.
~1,300 now; ~1.2 million by 2028 in the median scenario.
Source: The Neuron, AI Economics 101 (Mar 2026), token-to-salary conversion.
Usage
Tokens, prompts, pictures — the demand side of the meter.
Compute
Installed accelerators and the factories that feed them.
Energy
Watts, water, and carbon behind every completion.
Capital
What the buildout costs, and who is writing the checks.
Labor
People, premiums, and how many agents the grid can actually run.
Research
Papers, launches, and the pace of the frontier.
