Paid ads
Free Google Ads headline generator with character counts
By Charles Summers · Updated · Free, no signup
Short answer
This generates 15 responsive search ad headlines and 4 descriptions, then measures every single one against the limits Google currently documents: 30 characters for a headline, 90 for a description. Each asset shows its exact character count, a pass or fail flag, and an auto-trimmed rewrite if it is over. The 15 headlines are spread deliberately across six categories, keyword-match, benefit, offer, proof, objection-handling and call-to-action, and a pairwise overlap check flags any two that are too similar to be worth separate asset slots.
Use the google ads headline generator
What does this tool actually do?
This generates 15 responsive search ad headlines and 4 descriptions, then measures every single one against the limits Google currently documents: 30 characters for a headline, 90 for a description. Each asset shows its exact character count, a pass or fail flag, and an auto-trimmed rewrite if it is over.
It runs entirely in your browser. Nothing you type is sent to a server, no account is required, and there is no usage limit, because there is no cost per run to control.
The limit is a hard reject, not a truncation
A meta description that runs long gets cut off in the results page. A Google Ads headline that runs long does not get served at all. The editor refuses the asset at the point you save it, and the API returns a validation error rather than a trimmed string. That difference matters for how you write: with organic snippets you are optimising for what survives the cut, and with ad assets you are working to a fixed container that either accepts the string or does not.
Google counts characters, and spaces are characters. So are the pipes, hyphens and ampersands people use to squeeze two ideas into one line. Emoji and most symbols are disallowed in Search text assets under the editorial policies, which removes the usual trick for saving space. In languages written with double-width characters, Chinese, Japanese and Korean, the documented limits are halved, so a headline is effectively 15 characters and a description 45, which is why a direct translation of an English asset set almost never imports cleanly.
Two mechanisms can push an asset past the limit after you have written it, and both catch people out. Dynamic keyword insertion substitutes the matched search term at auction time, so a 22-character headline with an insertion tag in it can resolve to something much longer, at which point Google falls back to the default text you specified inside the tag. Ad customizers behave the same way, pulling a value from a feed that may be longer than the placeholder you counted. If you use either, count the longest possible substitution, not the one in front of you.
The current documented limits are 30 characters for each headline, 90 for each description, and 15 headline slots plus 4 description slots on a responsive search ad. Treat those as the values to check rather than as facts that will hold forever: Google has changed asset counts and added fields more than once, and the limits differ by campaign type on the same account.
Ad Strength grades variety, and variety is the thing that actually gets measured
Ad Strength is not a prediction of performance and Google has never claimed it is one. It is a scored checklist, and the inputs are legible: how many headline and description slots you filled, whether the text includes your keywords, and, most importantly here, how distinct the assets are from one another. That last input is why filling all 15 slots can leave you on Poor. Fifteen headlines that are permutations of the same six words score worse than eight genuinely different ones, because the grader is counting unique assets, not populated slots.
The reason this rule exists is mechanical rather than aesthetic. A responsive search ad is a combination engine: the system assembles headlines and descriptions into an ad at auction time and learns which arrangements earn clicks for which query. Its ability to learn anything depends on the combinations being meaningfully different. Feed it fifteen variations of the same claim and every combination is the same ad, so the system has nothing to test and no dimension along which to improve, regardless of how many slots you filled.
The practical consequence is that variety should be planned rather than hoped for. Decide the categories before you write, then write to fill them: match the query, state a benefit, state the offer, carry proof, answer the objection, and ask for the action. Six categories with two or three assets each fills the ad and guarantees the combination engine has real alternatives. This tool builds the set that way and then checks its own work with a token-overlap pass, because it is easy to write two "different" headlines that share every word that matters.
- Keyword-match assets exist to win the relevance half of the score and to earn the bolding that a query match produces in the served ad.
- Benefit assets say what changes for the customer. These carry the ad when the query is broad and the searcher has not decided what they want yet.
- Offer assets are the ones that move click-through most reliably, and the ones you must update the day the promotion ends.
- Proof assets need a real number or a real name behind them. An unsupported superlative can be disapproved under the misrepresentation policies, and in some markets superlatives require third-party verification.
- Objection-handling assets pre-answer the reason people bounce off the landing page: contract length, setup effort, migration, minimum spend.
- Call-to-action assets set expectation for what happens after the click, which is the cheapest way to stop paying for clicks that were never going to convert.
Pinning: what it costs, in combinations
Pinning locks an asset to a specific position, so a headline pinned to position 1 can only ever appear first. It is the correct tool for a genuine constraint: a legally required disclosure, a brand name your compliance team insists appears first, a regulated claim that must sit next to its qualifier. It is the wrong tool for a preference, and the cost of using it for a preference is arithmetic.
Unpinned, 15 headlines give the system an enormous number of possible three-headline arrangements to draw from, and every impression is an opportunity to learn which one suits the query. Pin one asset to position 1 and every served ad now starts with that single line, so the entire first position stops being a variable. Pin all three positions with one asset each and you have not built a responsive search ad at all, you have built one static ad with extra steps, and the reporting will show a single combination accumulating all the impressions.
If you must pin, pin more than one asset to the same position. Two or three headlines pinned to position 1 keeps a test running inside the constraint: the system still chooses between them, it simply cannot use any of them elsewhere. This is also how you avoid the Ad Strength downgrade that follows from heavy pinning, since the grader is reacting to the loss of combination variety rather than to pinning as such.
One more practical note. Google can serve two headlines instead of three, and it does so often on mobile where the space is narrower. If your ad only makes sense when all three headlines appear in a specific order, it is going to read badly a large share of the time. Write each headline so it stands on its own, and read the reporting on served combinations rather than assuming your intended arrangement is the one people are seeing.
The same copy does not fit Search, Performance Max and Display
Campaign type changes the asset specification, not just the audience. A responsive search ad takes up to 15 headlines at 30 characters and 4 descriptions at 90. A Performance Max asset group takes short headlines at 30, adds long headlines at 90 characters that appear on placements with room for them, and asks for a short description on a tighter limit alongside the standard ones. Responsive display ads take a 30-character headline, a 90-character long headline used when the placement is wide enough, descriptions that some placements drop entirely, and a business name field on its own short limit.
The behaviour differs as much as the fields do. In Search, your text competes on a page where every result is text, so specificity and query match do the work. In Performance Max and Display, your text is next to an image, on a page about something else, in front of someone who was not searching for anything. Copy that reads as a precise answer to a query reads as a non sequitur in a feed. The headline that wins on Search is frequently the exact one that fails everywhere else.
Because assets get mixed across placements automatically, every line has to survive being read alone, next to any image in the group, and without the line you assumed would sit above it. That is a real constraint on how you write, not a formatting detail: a headline that depends on a setup line will eventually be served without the setup. Write each asset as a complete statement, then let the system arrange them.
Numbers worth knowing
| Metric | Typical | What it means |
|---|---|---|
| Responsive search ad headline | 30 characters | Spaces and punctuation count. Currently documented value; halved for double-width languages such as Chinese, Japanese and Korean, so a translated set rarely imports without rewriting. |
| Responsive search ad description | 90 characters | Four description slots. Google typically serves two, and which two varies by device and query, so each one has to make sense on its own. |
| Asset slots on one RSA | 15 headlines, 4 descriptions | Filling every slot is not the goal. The grader counts distinct assets, so eight genuinely different headlines outscore fifteen rewordings of the same claim. |
| Performance Max long headline | 90 characters | An extra field Search does not have, used on placements with room for it. Asset groups also expect multiple images and a short description on a tighter limit. |
| Headlines served per impression | usually 2 or 3 | Google decides at auction and shows fewer on narrow screens. Any headline that only makes sense in a fixed sequence will be served out of that sequence regularly. |
Mistakes that quietly cost you results
- Filling all 15 headline slots with reworded versions of the same line
- The combination engine has nothing to test and Ad Strength reads it as low variety. Write to six fixed categories instead, and check that no two headlines share most of their meaningful words.
- Pinning your favourite headline to position 1 because it reads best
- That removes the entire first position from testing and usually triggers an Ad Strength downgrade. Pin only for compliance, and pin two or three assets to the same position so a test still runs inside the constraint.
- Counting characters in the version without the keyword insertion tag
- Dynamic keyword insertion and ad customizers resolve at auction, so the served string can be much longer than the one you counted, and Google falls back to your default text when it overflows. Count the longest possible substitution.
- Writing headlines that only work in a particular order
- Google serves two or three headlines from your set in an order it chooses, so a setup line and its punchline will regularly be separated or dropped. Every asset has to be a complete statement on its own.
- Reusing the Search asset set verbatim in Performance Max or Display
- The field specification is different, long headlines and business name have no Search equivalent, and the reading context is different. Query-matched copy reads as a non sequitur next to an image on an unrelated page.
What does the output look like?
This is the exact output the tool produces from the example inputs. It is generated by the same code that runs when you click the button, so what you see here is what you get.
Frequently asked questions
Why does an over-limit headline fail instead of getting cut off?
Because Google validates ad assets at the point you save them rather than at the point it renders them. The editor rejects a 34-character headline outright and the API returns a validation error, so nothing is served. That is the opposite of an organic search snippet, which is written into the page and then truncated in the results. It is why this tool shows a pass or fail flag and an auto-trimmed rewrite rather than a projected cutoff point.
Should I always fill all 15 headline slots?
Only if you have 15 genuinely different things to say. The Ad Strength grader responds to distinct assets, and the combination engine can only learn from combinations that differ, so fifteen rewordings of one claim score worse and test worse than eight lines that make different arguments. Fill the slots by working through the six categories this tool uses, and stop when you would have to start repeating yourself to continue.
Does the tool check that my headlines are different from each other?
Yes, and this is the part most character counters skip. After generating the set it tokenises every headline, drops common stop words, and compares each pair on shared meaningful words. Any pair sharing most of its vocabulary is flagged as a near-duplicate, because two assets that say the same thing occupy two slots and give the system one idea to test.
Do the proof headlines contain real statistics?
No, and deliberately so. Proof assets are written with bracketed placeholders such as [X] for you to replace with a figure you can actually defend, because an invented number in an ad is both a policy problem under the misrepresentation rules and a trust problem when a prospect checks it. Remember to recount characters after you substitute the real number, since a four-digit figure is longer than the placeholder.
Why do the character limits change when I pick Performance Max or Display?
Because the asset specification is different per campaign type. Performance Max asset groups add long headlines at 90 characters and a short description field on a tighter limit; responsive display ads add a long headline plus a business name field with its own short limit, and drop descriptions entirely on some placements. Selecting a campaign type here changes which fields get generated and which limits get applied to them.
Related free tools
- Facebook Ad Copy Generator Paid ads
- ROAS Break-Even Calculator Paid ads
- Ad Angle Generator Paid ads
- AI LinkedIn Hook Optimizer for job seekers Social
- AI LinkedIn Hook Optimizer for founders Social
Some links on this site are affiliate links, which means Hacking Demand may earn a commission if you buy through them at no extra cost to you. This does not influence which tools are listed. The tools on this page are free and have no affiliate relationship of any kind.