How much do ChatGPT ads cost? What we know (and don't)
OpenAI has not published a ChatGPT ad rate card. The scale of real advertiser activity is the closest thing to a benchmark we have.
There is no public rate card for ChatGPT ads. OpenAI runs advertising inside ChatGPT as a closed, invite-only pilot and has not released a CPM floor, a CPC benchmark, or a self-serve auction the way Google Ads or Meta Ads do. The honest answer is that the price is set behind closed doors, and the next best signal is the scale and concentration of real advertiser activity, tracked in our advertiser intelligence hub.
Even with no published pricing, 2,210 advertisers have already generated enough activity to fill a 73,720-placement leaderboard.
Why there is no public ChatGPT rate card
ChatGPT advertising is run as an invite-only program managed by OpenAI. Advertisers do not log into a public dashboard, pick keywords, and set a daily cap the way they would on a search or social platform. OpenAI controls which brands get in, what formats appear, and how each placement is priced. Until that changes, the answer to 'how much does a ChatGPT ad cost' is the same answer you would give for any private media buy: it depends on what you are buying, from whom, and at what scale.
The scale of activity that already exists
Who is spending the most on ChatGPT ads so far
We cannot infer a price from placement volume alone, but the concentration of activity at the top of the leaderboard is consistent with a high entry barrier. Mastercard, Cloudflare, Monday.com, Salesforce, and SAP are not experimenting at hobbyist scale. Each one is buying enough placements to rank in the top 10 across a 2,210-advertiser pool, and the top 10 advertisers combined hold roughly a quarter of all placements we logged, with Mastercard alone accounting for 4.73%.
- Others (2,200 advertisers)74%
- Mastercard5%
- Cloudflare4%
- Monday.com3%
- Salesforce2%
- Others12%
What factors drive the cost of a ChatGPT ad
Because there is no published rate card, the only way to estimate cost is to think about what the advertiser is actually bidding on. Four factors move the price of any given placement: the prompt that triggered it, the niche or vertical that prompt belongs to, the ad format OpenAI chooses to serve (a single recommendation versus a multi-product carousel), and the competitive intensity of that context. A query like 'best integration platform for go-to-market SaaS tools' triggers a different auction than 'where to find mid century modern furniture replicas,' and the brands competing in each context set their own bids. Whatever relevance, quality, or context signals OpenAI uses to rank ads can also push the price up or down, though the platform has not publicly disclosed which signals it weighs.
What prompts actually trigger these ads
The diversity of triggering prompts is the single biggest reason a single CPM number would mislead you. A finance question about tokenized real-world assets and a developer question about Jest TypeScript configuration are not the same auction. The first pulls in fintech and crypto advertisers; the second pulls in developer tooling and integration platforms. If you are budgeting a ChatGPT campaign, model spend by prompt intent, not by niche label.
Model spend by prompt intent, not by niche label.
We probe ChatGPT with realistic consumer prompts and capture the sponsored ad cards it returns — every creative, the triggering prompt, and the advertiser. Figures reflect our captured sample, not OpenAI's internal data. Explore the live ad library and market intelligence.