NEW New feature: Verify and block fake emailsLearn more
NEW New feature: Verify & block fake emails
Learn more
fraud blocker click fraud prevention logo
Sign in
Start Free Trial

ChatGPT’s New Ad Program: Pricing, Minimums & How to Sign Up

  • February 20, 2026
ChatGPT New Advertising Program
Key Facts at a Glance
  • Date Available: February 9, 2026 (U.S. Beta)
  • Minimum Spend: $200,000 to $250,000 for early partners
  • Bidding Format: Fixed CPM (Cost per 1,000 views) at roughly $60
  • Available Tracking: Total views and clicks only; No tracking pixels or conversion tags
  • How to Sign Up: https://openai.com/advertisers/

OpenAI has officially entered the advertising world. For years, ChatGPT was a tool you paid for through a subscription, but as of February 2026, the platform is now showing ads to millions of its free users.

If you are a marketing professional or a business owner concerned about ad fraud and “junk” traffic, this new platform requires a different strategy. It is not like Google or Facebook, and it comes with unique risks and high costs.

How ChatGPT Ads work

OpenAI is rolling out these ads slowly. Currently, they only appear for adult users in the United States who use the Free version of ChatGPT or the “Go” tier (the $8 per month plan). If you pay for the $20 Plus plan or use an Enterprise account at work, you will still see zero ads.

The ads themselves show up at the bottom of the screen after the AI has finished answering a question. They are clearly labeled as “Sponsored.” OpenAI has promised that the ads will never change the AI’s actual answer—the “organic” response stays independent from the paid content.

ChatGPT ad formats

There are three main ways your business can show up in ChatGPT:

  1. Sponsored Recommendations: Small text suggestions at the end of your conversation flow.
  2. Companion Display Ads: Visual boxes that include a product image and text.
  3. Native Content Cards: Larger, expandable boxes that pop up for specific shopping or travel queries.

Ad examples

Here is an example provided by ChatGPT (e are actively looking for ads in the wild and will update this section once we find any):

chatgpt-ads-example

Example ad flow provided by ChatGPT.

The pricing issue: why is it so expensive?

ChatGPT will be charging an incredibly high $60 CPM 🤯(cost per thousand impressions). 

This is about three times more expensive than ads on Meta (Facebook/Instagram). OpenAI is positioning this as “premium” space, similar to buying an ad during a live NFL game.

Comparison of CPM ad costs among different channels.

The idea is that when a user is talking to an AI, they are in a state of “active research.” They aren’t just scrolling mindlessly through a feed; they are looking for an answer. For brands, this means your ad is reaching someone at the exact moment they are making a decision. However, because the minimum spend is $200,000, this is currently a “big brands only” club.

Big concern: no tracking available

The key surprise with ChatGPT ads is the lack of tracking they are providing.

OpenAI does not allow you to use a tracking pixel so you cannot see which specific users clicked your ad or if they converted to a sale on your website. You only get a basic report showing how many people saw the ad and how many clicked it.

This creates a “Dark Traffic” problem. When someone clicks an ad in a ChatGPT conversation and lands on your website, your Google Analytics might show them as “Direct” traffic or “Unassigned” rather than a paid lead .

How to fix this: You must use UTM parameters on every link you submit to OpenAI. A UTM is a small snippet of code at the end of your URL that tells your analytics exactly where the user came from. For example: yourwebsite.com/?utm_source=chatgpt&utm_medium=paid. This is currently the only way to prove your ads are working. Read more about UTMs.

Brand safety: where your ads will (and won't) appear

OpenAI is being very strict about where ads show up to protect its reputation. Ads will not appear next to sensitive topics, including:

  • Health and medical advice.
  • Mental health discussions.
  • Politics and elections.

Additionally, ads are hidden from anyone the system thinks is under 18. For brands, this is a double-edged sword: your brand is safer from “bad” content, but you also can’t advertise in many popular categories.

How to sign up for ChatGPT ads and who gets in?

Currently, the OpenAI advertising program is not open to everyone. It is in a “managed beta” phase, which means you cannot simply sign up and launch a campaign today like you would on Google Ads.

The Application Process

  • Apply Directly – You can apply to run ads on ChatGPT here: https://openai.com/advertisers/
  • Get Invited – Most current advertisers were invited directly by OpenAI or reached through “managed service agreements” with major ad agencies.
  • Wait For Self-Serve: A self-serve platform for smaller businesses is expected later in 2026 or 2027, but no official date has been set.

Who is Getting Accepted?
OpenAI is very selective about who they allow into the beta. Here’s what we know so far:

  • Enterprise-Level Budgets: Because of the $200,000 minimum spend, the program is currently restricted to large, established brands with significant testing budgets. The large ad agency Omnicom has publicly stated that 30% of their clients have been approved for the beta. Their brands include Apple, PepsiCo, McDonald’s, Visa and Volkswagon.
  • Specific Industries: The initial rollout focuses on “high-consideration” categories where users do a lot of research. This includes Software (SaaS), Education, Professional Services, and E-commerce.
  • Prohibited Categories: To maintain trust, OpenAI is currently excluding ads related to Health, Mental Health, and Politics.

Strategic tips for advertisers

If your brand is is approved for ChatGPT advertising, consider these three rules:

  1. Think “Awareness,” Not “Direct Sales”: Because there is no conversion tracking, don’t expect to track every dollar spent back to a purchase. Treat this like a billboard or a TV spot—it’s for getting your name in front of people when they are thinking about your category.
  2. Monitor Your “Branded Search” Volume: Since many people will see your ad but not click it immediately, look for a “lift” in people searching for your brand name on Google while your ChatGPT campaign is running. This is a common sign that the ads are working .
  3. Use Dedicated Landing Pages: Since you can’t use pixels, create a specific page on your website that only ChatGPT ad traffic can reach. This makes it much easier to see if that traffic is actually converting into leads or sales.

Bottom line: we don't recommend ChatGPT ads

This is most certainly NOT for most advertisers and brands.

The excessive cost premium and the intentional lack of basic tracking features is designed for brands looking to capitalize on the publicity around these new ad units or simply those with enough money to throw around on ‘branding’ and little results. 

Thus, we cannot recommend any sales or performance-based marketers use this ad program today and hope OpenAI makes meaningful improvements in the future.

We’ll be waiting.

Frequently asked questions

OpenAI recommends keeping ads short and direct. Headlines should be between 50 and 80 characters. Descriptions should stay between 150 and 200 characters. Every claim or stat you use in the ad must be verifiable, as “hype language” (like calling yourself the “best”) may cause the ad to be rejected.

Yes. Advertisers have the ability to specify “exclusion” terms. For example, a premium software brand can exclude their ads from showing in conversations that mention “free alternatives” or “budget constraints” to avoid wasting their budget on users who won’t pay for their product.

This is a unique feature of ChatGPT’s ads. When a user sees your ad, they can tap a button to “Ask ChatGPT” about the product. This brings the ad details into the actual chat, allowing the user to ask follow-up questions, such as “What are the check-in times for this hotel?,” directly within the interface.

Facebook
X
LinkedIn

More from Fraud Blocker