OpenAI is changing how businesses pay for ads on ChatGPT, swapping manual bid controls for an automated system that sets its own cost targets. If you've ever run a Google or Meta ad campaign, the shift will feel familiar — because it's the same playbook, just arriving faster this time.
What happened
ChatGPT Ads is retiring its Max Bid and Max CPC (cost-per-click) settings, which let advertisers manually cap how much they'd pay for a click or impression. In their place, the platform is introducing "cost targets" — an average price point that OpenAI's bidding algorithm aims to hit across a campaign, rather than a hard ceiling an advertiser sets click by click.
The rollout isn't happening all at once. New campaigns will move to the cost target model first, existing campaigns get migrated on a separate, later timeline, and advertisers using the API will transition on their own schedule. That staggered approach gives OpenAI room to watch how the system performs on new, lower-stakes campaigns before forcing the change on accounts that are already spending real money.
The practical difference is control. Under Max CPC, an advertiser decides the absolute most they'll pay per click and the system respects that number, even if it means losing some auctions. Under cost targets, OpenAI's algorithm has latitude to bid above or below that target on individual auctions, as long as the campaign average lands near the goal. That flexibility is supposed to win more valuable impressions, but it also means less predictability on any single day.
This matters because ChatGPT Ads is still a young product. OpenAI has been building out its advertising business gradually, and bidding mechanics are one of the more consequential — and least visible — decisions a platform makes early on, since they shape how much businesses end up paying for the same visibility over time.
Why it matters
Google spent the better part of a decade nudging advertisers away from manual CPC bidding and toward automated strategies like Target CPA and Target ROAS, eventually making manual bidding harder to find and less competitive by design. Meta followed a similar arc with its cost-cap and bid-cap tools. In both cases, automation became the default not because advertisers demanded it, but because the platforms' algorithms perform better with more pricing freedom — and because it tends to increase overall ad spend on the platform.
ChatGPT Ads adopting the same structure this early suggests OpenAI is following that established model rather than inventing a new one. The genuinely new part here isn't the bidding mechanic itself — it's how quickly a brand-new ad platform is replicating a pattern that took its predecessors years to roll out.
What this means for small businesses
For a small business running or considering ChatGPT ads, the core trade-off is control for convenience. Cost targets can reduce the time spent fiddling with bids, but they also mean day-to-day costs become less predictable, especially while the algorithm is still learning from a smaller pool of campaign data than it would have on a mature platform like Google.
Budget caps become more important than bid caps in this environment. Since you can no longer set a hard per-click ceiling, the total daily or campaign budget is your main lever for limiting exposure. Businesses should also expect more volatility in cost-per-result during the first few weeks after their account transitions, simply because the algorithm needs data to find its footing.
It's also worth tracking whether average costs drift upward after the transition — a pattern seen on both Google and Meta once automated bidding became the norm, as platforms optimize for auction competitiveness rather than advertiser savings.
What to watch
Keep an eye on the specific migration date for existing campaigns, which OpenAI has not bundled with the new-campaign rollout. Watch for documentation changes on how cost targets are calculated and whether OpenAI publishes any guidance on acceptable variance from the target. API users should watch for a separate changelog entry, since that transition is being handled independently.
The bottom line
If you run ChatGPT ads, set a firm total budget before your account migrates, monitor cost-per-result daily for the first two weeks after the switch, and don't assume historical bid settings will carry over cleanly.