SEO title: Slug: chatgpt-ads-hit-1b-run-rate-openai-ad-audience-still-unknown
**
OpenAI says ChatGPT Ads reached a $1 billion annualized revenue run rate in less than 200 days, with tens of thousands of advertisers using the platform and more than one billion weekly active ChatGPT users. That is a current pace metric, not trailing recognized revenue, but it is large enough to establish advertising as an operating business rather than an optional future model.
The broadest denominator produces a deliberately conservative scale check. Dividing the $1 billion run rate by more than one billion weekly active users gives roughly $1 of annualized advertising revenue per total weekly active user. That is not ad ARPU: not every ChatGPT user is eligible to see ads, users have different activity levels, and OpenAI has not disclosed ad load or monetizable impressions. The calculation instead shows how little revenue per total user is required for the business to reach $1 billion when the audience is this large.
The more important figures are private. OpenAI has not disclosed the number of ad-eligible weekly users, impressions per user, price per thousand impressions, click-to-conversion economics, advertiser retention or the revenue share kept by buying partners. Those variables determine whether a $1 billion run rate is an early floor or already reflects the easiest inventory.
Distribution is widening. OpenAI's Ads Manager is available directly across India, Europe, the Middle East and North Africa, alongside technology and agency partners. Criteo says more than 2,000 brands were buying ChatGPT Ads through its platform by June and reports click-through rates two to three times comparable formats and more than 80% of ad-driven traffic coming from new customers. Those are Criteo channel results, not audited platform-wide benchmarks, but they provide the first external evidence that third-party buying access can add demand without OpenAI giving up the answer surface.
Muse is trying to own the next step after the answer
Meta's Muse, launched Sept. 8 in the United States, is a personal agent that runs in a dedicated secure virtual machine with its own browser. It can connect to services such as email and calendars, book appointments, fill forms and make purchases after user approval. Meta says Muse cannot see stored passwords or payment methods and does not share conversations or VM data with its advertising systems.
That creates a different commercial position. Advertising monetizes influence over discovery and consideration. A delegated agent that researches, negotiates and completes a purchase can sit closer to the transaction, where merchant access, attribution and payment economics could matter more over time.
Meta has not disclosed Muse revenue, transaction volume or retention, and says the service is free for most uses with subscription plans for heavier usage. So there is no basis yet for claiming that transaction execution is a better business than ChatGPT advertising. The strategic contrast is narrower: OpenAI has already monetized a very large intent surface, while Meta is testing whether users will delegate the action that follows intent.
For OpenAI, the next valuation-relevant disclosure is the ad-eligible denominator. It would turn the current $1-per-total-WAU scale check into an actual monetization measure and show whether growth is coming mainly from more users, more ad load or higher advertiser yield. For Meta, the equivalent proof is completed delegated transactions. The contest is no longer simply which assistant answers better; it is which product can control more of the path from intent to paid action without eroding the trust required to do so.
Sources
OpenAI - ChatGPT Ads reaches $1bn annualized run rate · OpenAI - New ways to buy ChatGPT Ads · Criteo - Integration update · Meta - Introducing Muse · Meta - Muse product page