OpenAI ChatGPT CPC Ads 2026: Pricing, Setup Guide & First Campaign Tutorial

What You'll Learn

• What OpenAI has publicly confirmed about ChatGPT Ads and its phased rollout.
• How CPM, CPC, and oCPC objectives change pricing and delivery decisions.
• How budgets, country targeting, platforms, context hints, and audiences fit into campaign setup.
• Which privacy, brand-safety, measurement, and eligibility checks belong before launch.

What ChatGPT Ads Are in 2026

OpenAI ChatGPT CPC Ads 2026 are part of a phased advertising program inside ChatGPT. The program places clearly labeled sponsored units below some responses when a relevant match exists. OpenAI says the ads run on separate systems from the chat model and do not influence the answers ChatGPT provides.

That separation is central to the product. An ad is a paid placement, not an answer and not an endorsement. OpenAI’s official testing announcement says advertisers do not receive users’ chats, chat history, memories, or personal details. The advertiser receives aggregate performance information such as views or clicks.

The legacy version of this article presented a fixed price card, a large minimum spend, a revenue target, and a universal self-serve launch as settled facts. The current official documentation supports a more precise explanation. ChatGPT Ads has CPM, CPC, and oCPC objectives, but the price paid depends on the objective, bid, auction, eligibility, and delivery.

Businesses can begin through the official OpenAI Ads landing page. The available account path, regions, daily budget, and review requirements can change as OpenAI expands the program. A guide should therefore explain the current documented controls and show how to verify access instead of promising that every advertiser can launch immediately.

For adjacent context on AI product workflows, see Current Affair’s analysis of the agent operating-system layer. Advertising inside a conversational product depends on the product’s user experience, privacy rules, and system boundaries.

Rollout Timeline and Regional Availability

OpenAI began testing ads in the United States on February 9, 2026. The initial test covered logged-in adult users on the Free and Go tiers. OpenAI said Plus, Pro, Business, Enterprise, and Education users would not have ads in that test.

On March 26, 2026, OpenAI said it planned to expand the pilot beyond the United States, beginning with Canada, Australia, and New Zealand. On May 7, 2026, the company said it planned pilots in the United Kingdom, Mexico, Brazil, Japan, and South Korea. An August 11, 2026 update said ChatGPT Ads had launched in those five additional markets and would continue expanding.

OpenAI’s August 18, 2026 announcement then said ChatGPT Ads would expand to 31 European markets the following week. It named Germany, France, Spain, Italy, Sweden, Norway, Denmark, the Netherlands, and Austria among the markets in that expansion. The announcement said European advertisers would initially use OpenAI’s Ads Solutions team, agency partners, and technology partners, with self-service Ads Manager access to follow later in the summer.

The public Help Center can show a different rollout snapshot from a product announcement. Its current general FAQ says ads are rolling out in the United States, Australia, New Zealand, and Canada, while the company announcements describe later additions and planned expansion. Treat the region as a live eligibility question. Check the account’s Ads Manager and the current OpenAI documentation before planning a campaign around a specific country.

Do not infer that an advertiser can target every market named in a press update on the same day. Campaign location pickers and account verification are more direct evidence of current access. OpenAI’s campaign help page also says the location catalog should be checked because supported locations can change.

Current Affair’s AI search optimization guide covers discovery in answer systems. ChatGPT Ads are paid placements and should not be described as organic citation or ranking.

How an Ad Appears and What It Does Not Change

OpenAI’s advertiser basics page describes an ad unit below a ChatGPT response. It can contain an advertiser name, favicon, headline, description, landing page, and image asset. The user should be able to distinguish the sponsored unit from the answer without interpreting the ad as part of the model output.

The general Help Center says a single ad unit may appear below a response when a relevant match exists during the test. It also says an ad may contain one or more items from an advertiser. The number of ads a user sees depends on the product state and the relevant match, so a campaign guide should not promise a fixed impression rate.

OpenAI says the ad system uses the current conversation context and intent. If personalized ads are enabled, selected signals from the broader ChatGPT experience can also be used. The advertiser does not receive the conversation that produced the match. The signal stays within ChatGPT and is used to decide relevance.

Ads are not eligible near sensitive or regulated topics such as health, mental health, or politics. OpenAI also says advertisers in sensitive or regulated verticals, including dating, health, financial services, and politics, are excluded at this time. These safeguards affect both user experience and campaign eligibility.

CPM, CPC, and oCPC Objectives

ChatGPT Ads currently supports three documented campaign objectives. CPM means cost per 1,000 impressions and is intended for reach and awareness. CPC means cost per click and is intended for engagement and traffic. oCPC means conversion-optimized cost per click and is intended for downstream actions after a click.

The objective changes both the way the campaign is optimized and the event that is billable. OpenAI’s campaign guide says CPM charges per 1,000 impressions, while CPC and oCPC charge per valid click. The oCPC objective optimizes toward a tracked conversion goal, but it does not charge per conversion.

Choose the objective before building the campaign structure. A reach campaign should use a message and audience that can be judged on exposure. A click campaign should use a landing page that can accept intent from a conversational context. A conversion-optimized campaign needs a reliable conversion event and enough measurement quality to distinguish a useful action from a simple page visit.

OpenAI says the campaign objective determines pricing and optimization. Its campaign guide also says that if a different objective is needed after creation, the advertiser may need to create a new campaign rather than changing the existing campaign into a different objective. Confirm the current interface before committing a large budget.

Current Affair’s AI tools guide is useful for product discovery, but its editorial comparisons should not be confused with a paid campaign objective or an OpenAI auction result.

What the $3 to $5 CPC Guidance Means

OpenAI’s advertiser basics guide recommends starting with a maximum CPC bid of $3 to $5 USD per click for CPC campaigns. That is a starting recommendation for a maximum bid. It is not a fixed rate card, a guaranteed clearing price, a promise of delivery, or a minimum spend.

The same guide says the system uses a relevance-weighted, second-price auction among eligible ads. The actual amount and delivery can depend on the eligible competition, the relevance of the ad and landing page, the campaign objective, the selected locations, the available conversations, and the bid strength indicated by Ads Manager.

This distinction corrects the old article’s claim that ChatGPT CPC Ads cost exactly $3 to $5 per click. An advertiser can use the range as a documented starting point, then judge real average CPC and delivery from the account’s reporting. The account should record the bid, spend, clicks, average CPC, and date range used for the comparison.

The recommendation also does not tell an advertiser whether a campaign will be profitable. A valid click can still produce no qualified lead or purchase. Landing-page relevance, product fit, conversion tracking, and customer value remain separate questions.

Do not copy an unofficial blog’s claim that a $50,000 or $200,000 minimum is required. The official campaign guide instead presents country-specific daily minimum budgets. Minimums can change with billing currency and account state, so the live campaign creation flow is the final check.

Budgets and Current Daily Minimums

OpenAI’s campaign guide supports either a campaign-total budget or a daily budget. A campaign-total budget is a total spending limit for the campaign duration, but it is not a pacing control. The page warns that spending may not be distributed evenly across the dates. A daily budget is a delivery target and actual spend can fluctuate above or below the selected amount on an individual day.

The following values are the country-specific daily minimums documented in the OpenAI Help Center page retrieved on August 21, 2026. They are not a universal minimum spend and they are not a guarantee that the campaign will serve.

A daily minimum is only a floor for campaign setup. It does not predict impressions, clicks, conversion rate, or return. A total budget can still be spent unevenly, and a daily target can fluctuate. Start with a controlled amount that can produce interpretable data, then adjust after the account records enough delivery.

Budget type can also be hard to reverse. OpenAI’s campaign guide says some changes from a campaign-total budget to a daily budget are not reversible. If the wrong budget type is selected, creating a new campaign may be required. Confirm this behavior in the current interface before launch.

Keep the budget record with the campaign objective, country, currency, bid, dates, and conversion event. This makes later comparisons meaningful and avoids presenting a budget minimum as if it were an average market price.

Creating a Campaign and Ad Group

The campaign is the container for a business goal. OpenAI’s current campaign guide says a campaign includes a title, objective, conversion event for conversion campaigns, budget, start and end dates, target countries, optional platforms, and optional custom audiences.

Start with one goal that can be measured. Use reach when the question is exposure. Use clicks when the question is traffic or engagement. Use oCPC when a tracked downstream action is the intended outcome. Write the conversion event in plain language so the person reviewing the report understands what counts.

Set dates that allow the campaign to collect meaningful data. A campaign can be paused or adjusted later, but changing multiple variables at once makes performance harder to interpret. Keep a record of the original objective and budget when a change is made.

Ad groups let an advertiser separate themes and context hints within the same campaign. OpenAI says context hints describe conversations, topics, or keywords where a product may be relevant. They guide matching but are not exact-match keywords and do not guarantee delivery in a particular conversation.

Keep each ad group focused enough that the landing page, creative, and context hints make sense together. If one group mixes unrelated products, a click or conversion report will be harder to interpret. Use clear names for region, product, objective, and message theme.

The official campaign creation guide says an advertiser should check account verification, billing, active campaign and ad status, campaign dates, and ad review when a campaign does not serve. These checks belong in the launch runbook.

Targeting, Context Hints, Platforms, and Audiences

ChatGPT Ads combines campaign controls with conversational relevance. A campaign can target countries. In the United States, OpenAI’s guide says supported states, designated market areas, and ZIP codes can be available where supported. The location picker is the source of truth for a particular account and date.

Platform targeting can include iOS App, Android App, or Web. OpenAI says Web includes desktop web and mobile web. Its current device reporting groups devices into Mobile and Desktop, with mobile web counted under Mobile. More detailed platform reporting may be added later, so do not assume the report will always split every surface.

Custom audiences can be included or excluded where the feature is available. Use them only when the account has an approved audience workflow and the campaign purpose justifies the added complexity. A custom audience setting does not override OpenAI’s user, safety, or regional eligibility rules.

Context relevance is different from search-keyword targeting. A person can describe a need in natural language without using the phrase an advertiser wrote in a hint. The system considers the conversation context, ad details, landing page, and other eligible signals. Design the ad and landing page for the problem people describe, not only for a list of keywords.

For context on how conversational systems process user intent, see Current Affair’s AI search optimization guide. Paid placement and organic answer relevance remain separate systems.

Creative, Landing Page, and Brand Safety Checks

OpenAI’s advertiser basics page says an ad unit can include an advertiser name, favicon, title, copy, landing page, and image asset. Prepare those elements as a single promise. The headline should match the page, the description should state a real benefit without a guarantee, and the image should support the same message.

A conversational context can be exploratory rather than transactional. A person may compare products, ask for constraints, or seek a basic explanation before choosing a provider. Landing pages should answer the question that brought the visitor there. Avoid sending every context to a generic home page when a more specific destination is available.

Brand safety affects campaign eligibility. OpenAI says ads are not eligible near sensitive or regulated topics, and it excludes advertisers in sensitive or regulated verticals such as health, financial services, politics, and dating at this time. Review the current OpenAI Ads Policies before submitting creative.

Users can hide or report an ad. A report can remove the ad from view and send it to OpenAI for review. Build a process for checking the actual ad rendering, landing page behavior, mobile layout, disclosures, and contact path before increasing budget.

Do not describe an ad as an OpenAI recommendation. OpenAI’s Help Center says paid placements do not mean OpenAI endorses the advertiser or its product. The distinction should also be clear in marketing copy and internal approval documents.

For a separate view of AI product selection, Current Affair’s AI tools comparison illustrates why an editorial recommendation and a paid placement need different labels.

Measurement, Conversion Events, and UTM Tracking

Ads Manager Beta reporting currently includes impressions, clicks, spend, click-through rate, average CPC, average CPM, and conversions. These fields let an advertiser compare delivery with the objective selected at campaign creation.

OpenAI’s advertiser basics guide says advertisers can set up conversion measurement in Ads Manager Beta. The official OpenAI Ads developer documentation should be checked for the current event and integration steps. Static UTM parameters can also be added to landing page URLs so existing analytics tools can identify traffic from ChatGPT Ads.

Choose the conversion event before choosing oCPC. A vague event such as “interest” is harder to test than a defined action such as a completed form, an approved signup, or a purchase confirmation. The event should have a stable definition and a review path for duplicate, delayed, or invalid signals.

Separate platform reporting from business reporting. Ads Manager may show valid clicks and conversions, while the destination analytics tool may show sessions, leads, or revenue under different attribution rules. Compare the definitions before calling the numbers inconsistent.

Do not publish a forecast from the $3 to $5 recommended bid. Measure the actual average CPC, conversion rate, qualified conversion rate, and customer value in the account. A real campaign result is evidence for that campaign and period only.

Privacy, Personalization, Age, and Safety Controls

OpenAI’s advertising principles state that ads should not influence ChatGPT answers, that conversations stay private from advertisers, and that people can control personalization. The Help Center says advertisers do not receive chats, chat history, memories, names, email addresses, precise location, IP addresses, or sensitive information.

When personalized ads are enabled, OpenAI may use past chats, memory, ad interactions, and the current thread to make ads more relevant. These signals remain inside ChatGPT. If personalization is turned off, OpenAI says ads can still use the current chat thread, but not other chat threads, ad history, or interests for selection.

Users can manage the experience in Ads controls. They can turn off personalization, clear ads data, dismiss an ad, report an ad, or ask why an ad is being shown. OpenAI says deleted ads data can be retained for up to 30 days before removal from its servers.

OpenAI says ads do not appear in accounts where a user says they are under 18 or where OpenAI predicts that the user is under 18. The Help Center also says ads do not appear in Temporary Chats and may not appear for logged-out users or after image generation during the described test.

Privacy is part of the campaign design, not a footnote. Document which event data goes to the advertiser’s analytics system, which signals stay inside ChatGPT, and which settings a user can change. Keep the campaign promise aligned with the actual controls.

Current Affair’s ChatGPT memory guide provides related context on memory controls. Memory in the user product does not mean advertisers receive memory data.

First Campaign Verification Checklist

Start at ads.openai.com and confirm that the account can create or request access to Ads Manager. Verify the business identity, billing currency, target country, supported platform, and current daily minimum displayed by the account. Do not rely on a third-party pricing card when the live product can show a newer value.

Choose the objective and write down the reason. For CPM, define the exposure goal. For CPC, define the valid click and landing-page action. For oCPC, configure a conversion event and confirm that the event fires once under the intended conditions. Record the maximum bid separately from the daily or campaign-total budget.

Prepare the ad name, favicon or logo, headline, copy, image, landing page, and context hints. Test the page on desktop web, mobile web, iOS, and Android where those surfaces are selected. Make sure the message does not imply OpenAI endorsement or a model answer influenced by payment.

Before launch, check that the campaign dates are correct, the ad has completed review, the billing state is active, and the selected locations are supported. Send a test visitor through the landing page and inspect the UTM parameters. Confirm that the privacy notice matches the analytics and conversion setup.

After launch, compare impressions, clicks, spend, average CPC, average CPM, CTR, and conversions using the same date range. Investigate delivery changes when the bid, context hints, location, platform, or creative changes. Keep the report with the campaign version so it can be audited later.

The accurate conclusion is narrower than the legacy claim. OpenAI ChatGPT CPC Ads 2026 has a documented CPC objective and a recommended maximum starting bid of $3 to $5, but it is not a fixed $3 to $5 price list. The program has CPM and oCPC options, country-specific daily minimums, changing regional access, privacy controls, and an Ads Manager workflow that must be verified in the live account.