/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

Sundar Pichai says Google is now processing 3.2 quadrillion tokens per month, up from 480T tokens per month a year ago and 9.7T tokens per month two years ago

and Doesn't Need You AnymoreSanuj Bhatia /Android Central:Google's new YouTube AI tools could make AI slop impossible to escapeAlistair Barr /Business Insider:Google's latest AI numbers are huge. Here are the stats CEO Sundar Pichai just dropped.Daniel Howley /Yahoo Finance:Google debuts biggest update to Search in 25 years, including AI agents

Axios Ina Fried

Context & Ripple Effects

Google’s reported monthly token processing has climbed from 9.7 trillion two years ago to 480 trillion a year ago, passed 1.3 quadrillion last summer, and now stands at 3.2 quadrillion. The related coverage ties that growth to broader deployment of Gemini-powered features across Google services.

That deployment is increasingly visible in Search: AI Mode reached all U.S. users, while Google has linked AI Overviews to query growth and reported 75 million daily AI Mode users. The token figure is therefore an operating-scale marker for a product shift already underway, not an isolated infrastructure statistic.

First-order effects

  • Google must provision and operate substantially more AI inference capacity as Gemini features serve more queries and tasks across its products.
  • The company gains a clearer scale signal for advertisers, developers, and partners evaluating whether Google’s AI features have moved beyond limited trials into mass-use services.

Second-order effects

  • Rival platforms face stronger pressure to publish comparable usage and deployment metrics, while cloud and AI-infrastructure providers benefit from sustained demand for capacity that can support high-volume inference.
  • As AI answers, agents, and generation features absorb more user activity, Google will have to keep proving that those experiences improve engagement without weakening the economics of its existing Search business.

Third-order effects

  • If token growth continues to track product rollout, competitive advantage in consumer AI will depend less on launching a model and more on financing, operating, and integrating inference at internet-service scale.
  • The pattern also raises the stakes around how major platforms measure and disclose AI value: token volume demonstrates utilization, but it does not by itself establish user benefit, revenue durability, or the quality of AI-generated content.

The trend: Consumer AI is shifting from model launches toward a contest to embed and run high-volume inference across incumbent platforms’ everyday products.

Discussion

  • @mweinbach Max Weinbach on x
    Top companies in Google Cloud are doing 1T tokens a day, if they move 80% of their workloads to Flash they can save ~$1B annually according to @sundarpichai
  • @alliekmiller Allie K. Miller on x
    Holy crap. And I don't say that lightly. Google is now processing 3.2 quadrillion tokens per month, up 7x from last year. That's a 3 with 15 zeroes after it. Actually, it's 3 and a 2 and 14 zeroes. #Google [image]
  • @edzitron Ed Zitron on x
    these guys will share literally any number other than “how much revenue we made on AI”
  • @officiallogank Logan Kilpatrick on x
    3.2 quadrillion tokens a month and still growing :) [image]
  • @kimmonismus @kimmonismus on x
    This is inane. Tokens processed at insane scale! [image]
  • @jamespmcleod.ca James McLeod on bluesky
    “Token” is an annoying and stupid word, and I kinda hate how it's permeated society through both AI and crypto.  [embedded post]
  • @google @google on x
    At last year's #GoogleIO, we were processing 480 trillion tokens a month across our surfaces. Now, we're processing over 3.2 *quadrillion* tokens a month. That's a 7x increase in just a year. These tokens represent problems being solved — by a user or a developer or a [image]