You can advertise on ChatGPT right now by signing up for the OpenAI Ads Manager, which launched in the United States on May 5, 2026, and is open to any business with a US billing address.
This is not a rumor or a beta invite. OpenAI opened self-serve access to its ad platform six weeks ago and has already hit $100 million in annualized revenue from it. If you run paid media for a business and you have not looked at this yet, you are probably going to hear about it from a competitor first.
This article covers how the platform works, what the ads actually look like, how targeting operates without cookies, what the early cost data shows, and whether this belongs in your 2026 media plan.
What is the OpenAI Ads Manager and how does it work?
The OpenAI Ads Manager is a self-serve advertising platform that lets businesses place sponsored content inside ChatGPT conversations. You create a campaign, set a budget, write your creative, and your ad appears as a sponsored response when a user's conversation matches your targeting criteria. It works on the web app, the iOS app, and the Android app.
OpenAI launched the platform on May 5, 2026. The early revenue numbers are striking: $100 million in annualized revenue in the first six weeks, with OpenAI projecting $2.5 billion in ad revenue for the full year. For context, Twitter took years to reach that kind of ad revenue. OpenAI is doing it on the back of 880 million monthly users and 2 billion queries per day.
Why OpenAI moved to ads now
OpenAI's subscription revenue from ChatGPT Plus, Team, and Enterprise is real but capped by willingness to pay. Advertising opens a second revenue line that scales with usage, not seat count. The ad-supported tier also lets OpenAI give free users more access in exchange for seeing sponsored content, which grows the overall user base and creates more ad inventory. This is the same model Google built its entire business on.
What does a ChatGPT ad actually look like?
A ChatGPT ad appears as a labeled sponsored response within the conversation. When a user asks something like "what's a good CRM for a small business," ChatGPT may respond with an organic answer followed by or integrated with a clearly labeled sponsored section that includes a product recommendation, a short description, and a link.
The ads are text-based and conversational in format. They do not look like banner ads. They read like a recommendation from within the chat. OpenAI has said the format will stay clearly labeled as sponsored to maintain user trust, and early reports from advertisers suggest users are engaging with them at higher rates than traditional display ads. The format rewards relevance over loudness.
What ad formats are available?
The current formats are:
- Sponsored text responses: A labeled recommendation within the ChatGPT reply, with a short description and a URL.
- Sponsored product cards: A structured block with product name, image (in supported contexts), price, and a click-through link.
- Sponsored follow-up prompts: A suggested question at the bottom of a response, labeled as sponsored, that leads users deeper into a product or service.
The platform is still adding formats. Video ads and audio placements within the voice mode of ChatGPT are expected later in 2026, but they are not yet in the self-serve system.
How does ChatGPT ad targeting work without cookies?
ChatGPT targets ads based on the live context of the conversation, not based on a cookie, a third-party profile, or browsing history. When a user asks about home insurance, the system reads the intent in that message and matches it to relevant advertisers in real time. No personal data from previous sessions is used for targeting.
This is a fundamentally different model from Google Ads or Meta Ads. Google Ads matches on keywords the user types. Meta Ads matches on a profile built from behavior across weeks or months. ChatGPT matches on what the user is thinking about right now. That real-time intent signal is new and advertisers are still figuring out how to use it well.
What targeting options does the Ads Manager give you?
Inside the Ads Manager, the main targeting levers are:
- Topic categories: You select broad interest areas (finance, health, travel, software, etc.) and OpenAI matches your ad to conversations in those categories.
- Keyword intent signals: You can add keywords that signal relevant conversations, similar to broad match in Google Ads.
- Geography: Country and region targeting, with city-level targeting available in the US.
- Device: Web vs. mobile app.
- User type: Free users vs. Plus subscribers (Plus users see fewer ads; free users see more).
There is no demographic targeting by age, gender, or income yet. There is no lookalike audience system like Meta's. The platform is early. Advertisers who have built their entire targeting strategy around demographic data will need to reframe how they think about audience selection here.
What do CPCs look like compared to Google Ads?
Early adopters are reporting CPCs significantly lower than comparable Google Search campaigns, in some categories 40 to 60 percent lower. This is typical of any new ad inventory. When demand from advertisers is low and supply is high, prices are low. Google CPCs in 2002 were a fraction of what they are today.
This window will not stay open forever. As more advertisers enter the platform and OpenAI tightens the auction, CPCs will rise. The businesses that move now, build their quality score equivalents, and learn what creative and targeting combinations work will have an advantage when the platform matures.
For reference, some advertisers in the legal, financial, and software verticals are reporting CPCs of $1.50 to $4.00 on ChatGPT for queries that would cost $8 to $20 on Google Search. That spread is the opportunity.
Running Google Ads alongside ChatGPT Ads? Ranksiege manages Google Ads campaigns from $199 per month, no contracts, live in 72 hours. If you want a professional audit of your current paid media setup before adding a new channel, get your free audit at free audit.
Who should actually advertise on ChatGPT in 2026?
The platform is a good fit for certain business types and a poor fit for others. Be honest about which category you fall into before committing budget.
Good fit:
- SaaS and software companies: ChatGPT users frequently ask for tool recommendations. "What's the best project management tool for a 10-person team?" is exactly the kind of query where a sponsored response can convert well.
- Professional services: Accountants, lawyers, consultants, and agencies get queries like "how do I find a good SEO agency in London" or "what does a CPA cost for a small business."
- E-commerce with strong product-market fit: If your product answers a question people ask ChatGPT, you belong here.
- B2B lead generation: Decision-makers are increasingly using ChatGPT for research. Reaching them during that research phase, before they visit a competitor's website, is valuable.
Poor fit right now:
- Businesses that depend on visual creative: The format is text-first. Fashion, food, and highly visual categories will wait for richer formats.
- Local service businesses with very small radius targeting: City-level targeting exists but neighborhood-level does not yet.
- Businesses with compliance restrictions on ad copy: Financial services and healthcare advertisers need to verify what copy restrictions apply before running.
In our work across campaigns for clients in the USA, UAE, and Australia, we have seen that channels tend to work best when the user's intent at the moment of seeing the ad matches the product closely. ChatGPT's contextual model does that well for informational and research-stage queries.
How do you set up your first ChatGPT ad campaign?
Setting up a campaign in the OpenAI Ads Manager follows a structure similar to Google Ads. Here are the steps:
1. Create an account at ads.openai.com with a US billing address.
2. Set campaign objective: Awareness, traffic, or conversions. Conversion tracking requires adding an OpenAI pixel to your website.
3. Set budget and schedule: Daily or lifetime budget. You can run continuous or date-bounded campaigns.
4. Choose targeting: Select topic categories, add keyword signals, set geography and device.
5. Write your ad: Headline (up to 50 characters), body text (up to 150 characters), display URL, and destination URL.
6. Submit for review: OpenAI reviews ads manually at this stage. Approval typically takes 24 to 48 hours.
7. Monitor in dashboard: The dashboard shows impressions, clicks, CTR, CPC, and conversions if you have the pixel installed.
The system is less complex than Google Ads at launch. There are no ad groups, no bidding strategy options beyond manual CPC and a basic target CPC automation, and no asset-based formats like Responsive Search Ads yet. This simplicity makes it easier to start but also limits optimization levers compared to mature platforms.
How should ChatGPT Ads fit into your broader media plan?
ChatGPT Ads should not replace Google Ads or Meta Ads right now. The platform is new, the measurement ecosystem is limited, and the audience, while large at 880 million users, skews toward more educated, tech-savvy users. Not every business's customer is in that group.
The right model for most businesses in 2026 is to treat ChatGPT Ads as a test allocation, a small percentage of a paid media budget (5 to 15 percent), running in parallel with established channels. Google Ads remains the highest-intent channel for most commercial searches. Meta Ads remains the strongest channel for visual creative, audience-based targeting, and retargeting.
What ChatGPT Ads adds is a new touchpoint in the research phase. Users asking ChatGPT for product recommendations or service advice are often at the consideration stage, not quite ready to buy but actively evaluating. If your business does well with top-of-funnel and mid-funnel content, this is a channel worth testing.
At Ranksiege, we manage multi-channel performance campaigns for clients across five countries. The teams that do best are the ones who test new channels deliberately: start small, measure what you can, optimize what works, and expand budget as confidence builds. We will be building ChatGPT Ads management into our service offering in Q3 2026 for clients who want a managed approach.
What mistakes are advertisers making on ChatGPT Ads right now?
The platform is new, which means the same errors that plagued early Google and Meta adopters are appearing again. The ones we are seeing most often:
- Copying Google Ads copy directly: Search ad copy is written for a user scanning a list. Chat ad copy is read inside a conversation. The tone needs to be more consultative and less salesy.
- Skipping conversion tracking: Running clicks without a pixel installed means you cannot measure ROI. Install the pixel before spending anything.
- Over-broad topic targeting: Selecting every relevant category at once burns budget fast. Start with one or two categories and expand.
- Ignoring the free-user vs. Plus-user split: Plus subscribers ($20/month) are a more engaged, higher-intent audience. If your product targets business users or early adopters, consider targeting Plus users specifically and accepting the smaller volume.
- Expecting Google-like data: The reporting is thin right now. If you need impression-share data, auction insights, or keyword-level quality scores, those do not exist yet. Plan for ambiguity.
- No creative variation: Test at least three versions of ad copy. The platform is young and no one has benchmarked what "good" looks like for your category.
The businesses that will win on ChatGPT Ads are the same ones that won early on Google and early on Meta: the ones who moved quickly, stayed methodical about testing, and did not chase perfection before they had data.
