Yes, you can advertise on ChatGPT in 2026. OpenAI began testing ads with US free-tier and ChatGPT Go users on January 16, 2026, launched fully on February 9, 2026, and opened its self-serve Advertise in ChatGPT platform, the OpenAI Ads Manager, on July 22, 2026, with CPC and CPM bidding, no minimum spend, and reported pricing around $3 to $5 per click. Ads now appear in roughly 49 percent of US ChatGPT replies as labeled boxes beneath the answer, and the program generated $100 million in annualized revenue within six weeks of the US test. What you still cannot buy, at any price, is the answer itself: citations and recommendations inside the response remain organic, which is why AEO and ChatGPT ads are two different investments solving two different problems.
How do ChatGPT ads actually work in 2026?
They are contextual answer-adjacent placements, not search keywords. The OpenAI Ads Manager targets users based on conversational context, the topic of the current chat, past chat history, and prior ad interactions, rather than bidded keywords in the Google Ads sense. When a free-tier or ChatGPT Go user asks something commercially relevant, an ad can appear in a clearly labeled, tinted box at the bottom of the response. OpenAI’s stated rule is that ads never influence the answer content above them: the model writes its response, then the ad system attaches a placement.
The rollout timeline matters for planning. Testing started January 16, 2026 for logged-in US adults on Free and the $8-per-month ChatGPT Go tier. Full launch came February 9. The early-access phase carried a $50,000 minimum that shut out small advertisers; the self-serve launch eliminated that minimum entirely. As of mid-2026, five markets are live: the US, Canada, the UK, Australia, and New Zealand, with ads appearing in about 26.5 percent of replies globally and 49 percent in the US. Paid tiers above Go remain ad-free, which means a meaningful slice of high-income users never sees a placement, a targeting gap worth remembering when you compare this channel to the organic answer layer we covered in how to get cited by ChatGPT.
What does it cost to advertise on ChatGPT?
Less than you would guess, for now. Reported pricing on the self-serve platform runs $3 to $5 per click with both CPC and CPM options and no minimum spend. Compare that to the auctions AI answers are displacing: legal keywords like “car accident lawyer near me” run $100 to $300 per click in Google Ads, and even mid-competition service queries routinely clear $20 to $50. Early AI ad inventory is underpriced relative to intent because the auction is young and most advertisers have not built conversational targeting competence yet.
Expect that gap to close. OpenAI booked $100 million annualized within six weeks of the US test, and every dollar of demonstrated advertiser demand pushes CPCs toward parity with established channels. The 2026 pattern mirrors early Facebook and early Google: a low-cost window, then repricing. If your category is served by ChatGPT ads today, testing early is rational; budgeting as if $3 clicks are permanent is not. For the organic side of the same budget conversation, our breakdown of how much GEO costs covers the retainer and tooling market.
Before you spend a dollar on AI ads, find out what ChatGPT already says about you for free. Run a free AI visibility audit and see which buyer prompts cite your brand organically, and which cite your competitors.
What can ChatGPT ads do that AEO cannot, and vice versa?
They solve different problems, and the split is worth stating precisely:
1. Ads buy presence; citations confer trust
An ad sits below the answer in a labeled box. A citation sits inside the answer as the substance of the recommendation. Users treat them accordingly: the answer is advice, the ad is an ad. For high-trust purchases, legal, medical, financial, B2B software, the recommendation layer drives the decision, which is why AI-referred organic traffic converts at rates far above paid or classic organic, a pattern we documented in why AI traffic converts better.
2. Ads stop when spend stops; citations compound
Every ad impression is rented. A citation earned through structured content, entity consistency, and directory presence persists across thousands of future conversations at zero marginal cost, and tends to reinforce itself as engines re-retrieve sources they have already cited.
3. Ads reach free-tier users only; citations reach everyone
Plus, Pro, Team, and Enterprise users see no ads, and those tiers skew toward exactly the high-income professional buyers most service businesses want. Citations appear at every tier, including inside the paid accounts your best customers use.
4. Ads are available now in five countries; citations are global
If your market is outside the US, Canada, UK, Australia, and New Zealand, ads are not yet an option. The organic answer layer has no geography gate.
How do you actually set up a ChatGPT ad campaign?
Through the OpenAI Ads Manager, and the mechanics will feel familiar even though the targeting logic is new. You create an account, define campaign objectives and budget, and build targeting around conversational context categories rather than keyword lists: the system matches your ads to conversation topics, informed by chat history and past ad engagement. Creative is text-forward and answer-adjacent; the placement format is a labeled box, not an interruption. Measurement runs through the platform’s reporting plus your own UTM discipline, and if you are serious about attribution you should already be tracking AI-sourced sessions the way we outlined in how to track ChatGPT traffic in GA4.
Policy review is the step early advertisers most often skip. OpenAI applies category restrictions familiar from other ad platforms, with tighter handling in sensitive verticals: parts of health, legal services, and financial products face limits or additional review, and political advertising is excluded. Service businesses in regulated categories should read the current policy list before building campaigns, because approval standards are still moving quarter to quarter as the platform matures, and a rejected campaign after creative production is an avoidable cost.
Three practical rules from the early cohort. First, match the ad to conversation intent, not to your product taxonomy; the users seeing your placement are mid-conversation about a problem, not browsing. Second, send clicks to direct-answer landing pages, because ChatGPT users arrive expecting the page to continue the conversation’s specificity. Third, watch frequency: conversational context targeting can re-serve aggressively inside long chats, and early advertisers report fatigue effects the platform’s controls are still catching up to.
Should your marketing budget go to ChatGPT ads or AEO first?
AEO first, ads as acceleration, for one structural reason: the ad platform monetizes the audience, but the answer layer allocates the demand. When 49 percent of US replies carry an ad, 100 percent of replies carry an answer, and the answer names someone. If that someone is your competitor, no adjacent ad box repairs the loss, because the user asked a question and the answer is what they came for.
There is also a measurement asymmetry worth pricing in. Ad performance arrives in a dashboard: impressions, clicks, spend, conversions, all attributed by the platform selling you the inventory. Citation performance requires your own instrumentation, prompt tracking, referral analysis, share-of-voice measurement, which is why many teams overweight ads simply because ads are easier to report on. That is a reporting bias, not a performance finding. The businesses measuring both channels properly consistently find that cited queries produce fewer clicks but far higher close rates, because the engine has already done the qualifying before the visitor arrives.
The efficient sequence for most service businesses: establish organic citation coverage on your buyer queries, which also builds the entity and content foundation engines keep rewarding, then layer ads on the commercial prompts where you want incremental presence or where citations are still forming. Brands that run ads with no citation footprint pay to appear next to answers recommending other companies, which is the most expensive way to learn what getting your brand mentioned by AI would have cost.
FAQ: advertising on ChatGPT in 2026
Can anyone advertise on ChatGPT now?
Any advertiser in a live market, yes. The self-serve OpenAI Ads Manager, opened July 22, 2026, removed the $50,000 minimum from the early-access phase and offers CPC and CPM bidding with no spend floor. Live markets as of mid-2026 are the US, Canada, the UK, Australia, and New Zealand. Standard ad policy categories apply, and some verticals face restrictions, including parts of legal and health.
How much do ChatGPT ads cost in 2026?
Reported pricing runs $3 to $5 per click on the self-serve platform, with CPM options for awareness buys and no minimum spend. That is a fraction of comparable high-intent search CPCs, which reach $100 to $300 in categories like legal. Pricing reflects a young auction; OpenAI’s $100 million annualized revenue in six weeks of testing signals demand that will push rates up as inventory matures.
Do ChatGPT ads affect what the AI actually recommends?
No. OpenAI states ads appear in labeled boxes below responses and never influence answer content, and no advertiser can purchase a citation or recommendation inside the response. The recommendation layer remains organic, driven by retrieval, source trust, and content structure. That separation is the entire reason AEO exists as a discipline independent of AI advertising.
Who sees ads in ChatGPT?
Logged-in adult users on the Free tier and the $8-per-month ChatGPT Go tier in live markets. Plus, Pro, Team, and Enterprise subscribers see no ads. Ads currently appear in roughly 49 percent of US replies and 26.5 percent globally, weighted toward commercially relevant conversations. That leaves the highest-spending subscriber tiers reachable only through organic citations.
Can you buy your way into AI answers on other platforms?
Perplexity has experimented with sponsored follow-up questions, Microsoft Copilot inherits Bing’s ad system, and Google blends AI Overviews with its existing ad stack, but on every major platform the generated answer itself remains organic. No engine currently sells placement inside the recommendation, which keeps citation-building the only path into the answer layer on all of them.
Is early ChatGPT advertising worth testing?
If your buyers are US free-tier consumers and your category is approved, yes, cautiously: $3 to $5 clicks against conversational intent is cheap discovery, and early-auction inefficiency favors testers. Treat it as a complement to citation coverage, not a substitute. Set hard budget caps, instrument attribution before launch, and expect repricing as the auction matures.
The bottom line on ChatGPT ads
ChatGPT advertising is real, self-serve, and cheap in 2026, and it changes nothing about the fundamental economics of AI search: the answer is the product, the answer is organic, and the answer names whoever built the strongest citation footprint. Ads are a rational accelerant for brands that already show up in the response, and an expensive adjacency for brands that do not. Sequence accordingly, spend against instrumented attribution, and remember that the $3 click era is a window, not a price.
Want to know whether you are ad-adjacent or answer-included? Get your free AI visibility audit and see exactly where your brand stands in ChatGPT’s organic answers before you fund the ad box next to them.
Tagged