OpenAI Introduces ChatGPT Visual Ads During Image Generation

On October 5, 2026, OpenAI announced a new advertising format, ChatGPT Visual Ads. Its distinguishing feature is the placement of visual advertisements directly while images are being generated in ChatGPT.

The company announced that it would test the format in the United States during October with a limited group of advertisers. The ads are intended for users on the Free and Go plans.

What Are ChatGPT Visual Ads?

Visual Ads are an advertising format that allows businesses to showcase products and services using multiple images within a single ad unit.

For example, a user asks ChatGPT to create an illustration of an apartment interior. While the system generates the image, the user may see an advertisement from a furniture brand featuring photos of home furnishings.

The ad creative can showcase individual products, a product assortment, and use cases.

How the New Advertising Format Works

1. A user submits an image-generation request in ChatGPT.
2. While ChatGPT creates the image, the interface may display a visual advertisement.
3. The advertisement contains multiple images from the same advertiser and is clearly labeled as sponsored content.
4. The user can click the ad to visit the advertiser’s website.
5. Once image generation is complete, the ad unit disappears.

The ad does not become part of the generated image. The advertiser cannot alter the generated content or influence ChatGPT’s responses.

How Visual Ads Differ From Existing ChatGPT Advertising

The main difference between Visual Ads and standard ChatGPT Ads is where and when the ad appears. Standard ads are displayed below completed text responses in ChatGPT, whereas Visual Ads appear directly during image generation and disappear after it is finished.

Using several images is not unique in itself: standard campaigns already support the Multiple Images format. Visual Ads therefore represent a new advertising placement, rather than simply a new type of image creative.

Who Can Access ChatGPT Visual Ads?

The initial test conditions are:

• Geography: United States.
• Users: ChatGPT Free and Go.
• Advertisers: a limited group of participants.
• Placement: during image generation.
• Creative: multiple images from one brand.
• Interaction: a click through to the advertiser’s website.
• Status: testing beginning in October 2026.

Plus, Pro, Business, and Enterprise subscribers are not included in the audience eligible to see these ads.

How Ad Targeting Works

ChatGPT Ads uses the context of a conversation and inferred user intent to assess ad relevance.

The system may consider the current conversation, the ad copy, the landing page, and Context Hints supplied by the advertiser. When ad personalization is enabled, additional signals from the user’s interactions with ChatGPT may also be used.

This means ads are not necessarily tied to individual keywords as they are in traditional search advertising.

However, OpenAI has not disclosed a separate targeting algorithm specifically for Visual Ads. It would therefore be inaccurate to assume that every mechanism used in standard ChatGPT advertising applies to the new format without modification.

When Will Visual Ads Become Available to All Advertisers?

In its October announcement, OpenAI did not provide a date for the general availability of Visual Ads or publish separate technical specifications for ad images in this experimental placement.

The company also did not release public CTR, CPC, CPA, or ROAS performance benchmarks for the new format.

As of early November 2026, Visual Ads remain an experimental advertising placement offered to a limited group of advertisers. Standard ChatGPT Ads and the Multiple Images format should be evaluated separately from this test.