Daily Budget or Total Campaign Budget on ChatGPT Ads
By Keshav Parsai 7 min read
In brief
Choose a daily budget on ChatGPT Ads, because a campaign-total budget hides underdelivery inside its own pacing and the switch away from it is one-way.
Last verified: 2 October 2026 | Version: 1.1 | Next scheduled review: 2 November 2026
Daily, for a first campaign, and OpenAI says so itself. Its campaign documentation states that if you are new to the platform it recommends starting with a daily budget so you can monitor delivery and performance before committing to a campaign-total budget.
There is one more reason it does not state, and it is the better one.
What each mode actually does
A daily budget is the average amount the campaign spends per day across a Sunday-to-Saturday budget week in your account's time zone. Actual spend on any given day may run above or below the figure you set, up to twice it on a single day, and the week cannot exceed seven times it. The minimum is set by billing currency: 25 USD, 15 EUR, 15 GBP and so on across 30 currencies OpenAI publishes. On fixed bidding, the daily budget must also be at least your highest effective ad-group bid.
A campaign-total budget is the total amount you want to spend across the full duration of the campaign. It is a harder ceiling on the whole flight rather than a soft average on each day. OpenAI paces it against your remaining budget and schedule: up to the end date, capped at 365 days, or a default 60 days if you set no end date.
Both are set at campaign level, alongside objective, dates, locations and custom audiences.
The setting that cannot be undone
OpenAI documents one constraint that deserves more attention than it gets: if you change a campaign-total budget to a daily budget, you cannot change that campaign back to a campaign-total budget.
The switch runs one way only. So the choice is not symmetrical, and it should not be treated as a preference you can revisit. Starting on daily leaves you with a decision you can still make later, at the cost of rebuilding the campaign if you want total. Starting on total leaves you with a door that closes the first time you walk through it.
When two options are otherwise close, take the reversible one.
Why a total budget makes underdelivery hard to diagnose
This is the part that matters for a fixed-length test, and it is specific to a channel this young.
The most common failure on ChatGPT Ads is not overspending. It is not spending. New accounts frequently find that a campaign does not exhaust its budget, because context hints match narrowly, the auction is relevance-weighted, and there is no query volume data to tell you in advance whether the situations you described occur often enough to serve against.
Now consider how each budget mode reports that failure.
With a daily budget, underdelivery is legible on day two. You set 100 USD a day, the campaign spent 34, and the gap is a number you can read without doing any arithmetic. You know immediately that the problem is match volume or bid, not budget.
With a campaign-total budget, the same failure hides. A 3,000 USD total across 30 days is being paced by the platform against a duration, so slow days do not look like slow days. They look like pacing. OpenAI's Budget Pacing documentation confirms the mechanism: delivery slows when a campaign is ahead of pace and continues when it is behind, so it has a chance to catch up. A campaign that is behind pace because it genuinely cannot find matches looks, from the outside, like one the system is still expecting to catch up. You cannot tell the two apart until the flight is nearly over.
By the time the gap is undeniable, you are usually two thirds of the way through a fixed-length test with most of the budget unspent and no time left to fix the bid or the hints.
When a total budget is the right choice
It is not a trap. It is a mismatch for diagnosis, and a good fit for commitment.
Use a campaign-total budget when the money is genuinely fixed and the flight is genuinely bounded: a conference, a product launch window, a quarter-end push where finance has approved a number rather than a rate. In those cases the hard ceiling is the feature, and you have already accepted that you will not be tuning mid-flight.
Use it after the channel has proved it can spend, not before. The order matters. Prove delivery on daily, then commit on total if the flight shape calls for it, and accept that the campaign you commit with is a new one.
The honest limit of this advice
The first version of this page said campaign-total pacing was undocumented, and that the argument rested on how pacing systems generally behave. OpenAI has since published how it works: catch-up when behind, slow-down when ahead, against the end date or a 60-day default. That supports the argument. What remains undocumented is how aggressively the catch-up runs late in a flight, so a total budget that is behind pace may still spend unevenly at the end.
One more practical point from the same documentation: set an end date. Without one, a total budget paces over 60 days whatever you had in mind.
What we cannot tell you
- How hard campaign-total catch-up runs late in a flight. OpenAI documents the catch-up and slow-down mechanism but not its limits.
- Whether the budget mode affects auction behaviour or delivery volume. Undocumented.
- What proportion of new campaigns underdeliver. Unpublished by OpenAI, and no cross-advertiser benchmarks exist.
- Whether the one-way restriction applies in reverse under any circumstance. OpenAI states the total-to-daily change cannot be undone and does not describe exceptions.
Quick answers
Should I use a daily or total budget on ChatGPT Ads? Daily for a first campaign. OpenAI recommends it directly, and it makes underdelivery visible on day two rather than in week three.
What is the minimum daily budget? It depends on billing currency: 25 USD, 15 EUR, 15 GBP, 725 INR, 4.00 SGD and so on. See every currency on our country directory. There is no minimum lifetime commitment.
Will a daily budget spend exactly that amount each day? No. It is an average across a Sunday-to-Saturday week. A single day can spend up to twice the figure, and the week cannot exceed seven times it.
Can I switch between budget types? Only once, and only in one direction. Changing a campaign-total budget to a daily budget is permanent for that campaign.
When does a total budget make sense? When the flight is genuinely bounded and the money is genuinely fixed, such as an event or a launch window, and after the channel has already proved it can spend.
Does a total budget mean I will spend the whole amount? No. It is a ceiling, not a commitment, and underdelivery is common on this channel.
Sources
| Claim | Source | Tier |
|---|---|---|
| OpenAI recommends starting with a daily budget for new advertisers | OpenAI Help Center, Create Campaigns for ChatGPT Ads, 2026 | Confirmed, primary |
| Daily budget averages over a Sunday-to-Saturday week; up to 2x on a day, 7x over a week | OpenAI Help Center, Daily Budgets, 2 Oct 2026 | Confirmed, primary |
| Campaign-total budget sets the total spend across the full campaign duration | OpenAI Help Center, Create Campaigns for ChatGPT Ads, 2026 | Confirmed, primary |
| Changing a campaign-total budget to a daily budget cannot be reversed | OpenAI Help Center, Create Campaigns for ChatGPT Ads, 2026 | Confirmed, primary |
| Minimum daily budget by billing currency, 30 currencies | OpenAI Help Center, Create Campaigns for ChatGPT Ads, 2 Oct 2026 | Confirmed, primary |
| Daily budgets were added to Ads Manager Beta alongside geo-targeting controls | Search Engine Land, 22 May 2026 | Reported |
| Campaign-total budgets pace against remaining budget and schedule, slow when ahead, catch up when behind; 365-day maximum, 60-day default | OpenAI Help Center, Budget Pacing, 2 Oct 2026 | Confirmed, primary |
Related reading
- How Many Ad Groups Should One ChatGPT Campaign Have?
- Budget Calculator
- What ChatGPT Ads Cost
- How the ChatGPT Ads Auction Works
- ChatGPT Ads Troubleshooting: Why Your Campaign Is Not Spending
Changelog
2 October 2026, v1.1. OpenAI published campaign-total pacing behaviour, so the undocumented-mechanism caveat is replaced with the documented one. Added the 2x daily and 7x weekly limits, the Sunday-to-Saturday budget week, the fixed-bid constraint and minimums by billing currency.
12 September 2026, v1.0. First publication. Records the one-way total-to-daily switch and argues the choice on diagnosability rather than on spend control.
Field kit
Tools
Site
OpenAI (primary)
Share this
Keshav studies how AI systems retrieve, verify, and cite brand information. At InPromptAds, he leads source research and turns platform documentation into practical guidance for advertisers.
