Users Can Opt Out of ChatGPT Ads. Nobody Has Published What That Does to Your Reach
By The InPromptAds Team 6 min read
Last verified 15 August 2026. Version 1.0. Next scheduled review 15 September 2026.
Search for ChatGPT ads opt-out and you get dozens of articles written for users, explaining how to stop seeing ads. Every one of them is written from the reader's side of the screen.
Nobody has written the other side. Every user who opts out leaves the pool an advertiser is bidding on, and OpenAI has not published how many have.
That is not a minor gap. It is the denominator.
The audience is a residual, not a population
The number quoted in almost every article about this channel is total ChatGPT users. That number is not the addressable audience and it is not close.
| Start with | Total ChatGPT users |
|---|---|
| Subtract | Users on Plus, Pro or any Business plan. OpenAI states ads are not shown to them |
| Subtract | Users OpenAI predicts are under 18 |
| Subtract | Free users who take the Ads-Free option |
| Subtract | Sessions in conversations about personal health, mental health or politics, and other sensitive user journeys, where no ads run regardless of who the user is |
| Result | The ad-eligible pool |
Three of those four subtractions are documented. The size of none of them is published.
There is one circulating figure worth handling carefully: a pilot-period report that fewer than 20% of eligible users were being shown ads on a given day. It is undated on most pages that quote it, it describes delivery rather than eligibility, and it comes from a period that has since ended. It is not a substitute for the numbers above.
The exclusion nobody costs
The Ads-Free option lets a Free user leave the ad-eligible pool without paying, in exchange for lower message limits.
Think about what that selects for. The user who notices the option, understands the trade, and decides that fewer messages is worth an ad-free experience is a heavier, more deliberate, more product-literate user. That is not a random slice of the free base. It is plausibly the most valuable slice of it for most advertisers.
And there is no published take-up rate. Not from OpenAI, not from any third party, not in any of the agency guides selling this channel. Which means every media plan for ChatGPT ads currently contains an unstated assumption that the opt-out rate is small, and nobody has checked.
We are not claiming the rate is high. We are pointing out that nobody knows, including us, and that a channel where the addressable audience cannot be sized is a channel where reach forecasts are decoration.
Why this compounds with the plan exclusion
The Business plan exclusion and the opt-out exclusion pull in the same direction, and they hit the same kind of user.
The people most likely to sit on a Plus, Pro or Business plan are heavy users, professionals, and people whose employers pay for their tools. The people most likely to take an Ads-Free trade are heavy users who care enough to change a setting. The two exclusions are not independent, and they both remove the engaged end of the distribution.
What remains is a lighter, more casual, more consumer-shaped audience. That is a real and reachable audience, and it maps well onto OpenAI's approved categories: household goods, local services, travel and experiences, digital products and education.
It maps badly onto anything sold to a professional at work, which is the case we have argued separately in who should not buy ChatGPT ads yet.
What an advertiser should do about it
Stop quoting total user counts in planning documents. Any deck citing ChatGPT's total user base as the reach for an ad campaign is off by an unknown but certainly large factor, in a direction nobody can currently quantify.
Plan on impressions, not on audience. CPMs in this channel are priced on views rather than clickthrough, which means the platform is already selling you delivery rather than reach. Model the buy the same way and the sizing problem stops being load-bearing.
Treat the opt-out rate as a risk line, not a footnote. If your case for the channel would break at a 20% opt-out rate, you do not have a case yet, because you cannot rule that out.
Ask your rep, and tell people the answer. Advertisers in the pilot are better positioned than we are to ask OpenAI directly what share of the free base has taken the Ads-Free option. If someone gets an answer, that number is worth more to this category than another eligibility guide.
What we cannot tell you
The Ads-Free take-up rate. Nobody has published it. This is the point of the page.
Total ad-eligible users in any market. The exclusions are documented and unsized, so the residual cannot be calculated from public information.
Whether opt-out skews the way we have argued. The selection logic is reasoning, not measurement. A heavier user taking the trade is plausible and unproven, and we would revise this section if data appeared showing the opposite.
Whether OpenAI intends to publish any of this. No indication either way.
Quick answers
Can ChatGPT users opt out of ads? Yes. Free users can take an Ads-Free option, which removes them from the ad-eligible pool in exchange for lower message limits. Users on Plus, Pro or any Business plan are not shown ads at all.
How many ChatGPT users see ads? Not published. Ads are excluded for Plus, Pro and Business plan users, for users OpenAI predicts are under 18, and for Free users who take the Ads-Free option, and OpenAI has published the size of none of those exclusions.
Does opting out of ChatGPT ads cost anything? Not in money. The Ads-Free option trades ad-free usage for lower message limits on the Free tier.
Should I use ChatGPT's total user count to forecast ad reach? No. The ad-eligible audience is that number minus four documented exclusions whose sizes are unpublished. Forecasts built on the total will overstate reach by an unknown factor.
Does the opt-out affect which users I reach, not just how many? Probably, though this is reasoning rather than measured data. Users who take the trade are likely to be heavier and more deliberate, and the plan-based exclusion removes a similar group, so both exclusions reduce the engaged end of the audience.
Sources
| Claim | Source | Tier |
|---|---|---|
| Ads not shown to Plus, Pro or Business plan users, or predicted under-18s | OpenAI Help Center, Ads in ChatGPT, updated 30 July 2026 | Confirmed, primary |
| Free users can take an Ads-Free option with lower message limits | OpenAI | Confirmed, primary |
| Take-up rate for the Ads-Free option | Not published by OpenAI or any third party | Absent |
| No ads beside personal health, mental health or political conversations, or sensitive user journeys | OpenAI ad policies v1.3, 15 July 2026 | Confirmed, primary |
| Ads run on Free and Go tiers | OpenAI communication to users, reported 14 Aug 2026 | Reported |
| Fewer than 20% of eligible users shown ads daily | Pilot-period reporting, undated on most sources | Reported, stale |
| CPMs priced on views rather than clickthrough | eMarketer, 10 Feb 2026 | Reported |
| Approved categories: household and consumer goods, local services, travel and experiences, digital products, education | OpenAI ad policies v1.3 | Confirmed, primary |
Related reading
- Who should not buy ChatGPT ads yet — the six disqualifiers, including the Business plan exclusion
- How to measure ChatGPT ads — what the platform reports and what it does not
- ChatGPT ads changelog — dated record of documentation changes
Changelog
15 August 2026. v1.0 published.
Field kit
Tools
Site
OpenAI (primary)
The InPromptAds Team
Writes about AI-native advertising and the shift from search to conversation.
