Generate eight titles, check duplicates against recent public uploads, or review an existing title for clarity, specificity and fit with your supplied video facts. Editorial advice, not a click prediction.
One or two sentences: what actually happens in the video, and who it is for. Specifics in, specifics out.
Each option frames the same video differently, so you are choosing between approaches rather than between synonyms.
Copy the closest option, swap in your real numbers and names, and read it out loud before it goes on the video.
Most title generators stop at giving you options. This one adds a second step: paste your channel's public handle and it compares the eight titles against the last 50 videos already on that channel, so a title you have effectively used before gets flagged before you publish it again.
The check reads the same channel page anyone can open. It never asks you to connect an account and never reads private analytics, so it works on a channel you do not own as easily as on your own.
When a title repeats something, you get the old video's title, its view count, how long ago it went up and a link to open it — enough to decide whether to change the new title or retitle the old video.
A gaming channel whose every title says Minecraft would otherwise look like one long repeat. Words that appear across most of your titles are treated as your channel's background and excluded, so only the distinguishing words count toward a match.
The comparison covers the last 50 public uploads — not the whole channel. If the handle has no public videos, or too few to tell a repeat from a coincidence, the page says so rather than guessing.
The generator works from the sentence you type and nothing else. Kabo's Trend Radar previews a planned evidence-led workflow, but it does not currently watch your niche or report which angles are landing each week.
Aim for roughly 60 characters if you want the whole title readable. YouTube allows 100, but search results, suggested videos, and mobile home feeds all cut titles off well before that. Put the part that earns the click in the first few words, and treat anything past 60 characters as bonus context rather than something viewers will see.
A clear promise plus one specific detail — a number, a constraint, a result, a stake. Titles that only name the subject give a viewer nothing to decide with. It also helps when the title and the thumbnail say two different things that add up to one idea, instead of repeating each other word for word.
Front-load the keyword when it is also the hook — it survives truncation and it matches what people typed. Do not force it when the result reads like a filing label; a title that sounds like a human wrote it will beat a keyword-first title that nobody wants to click on.
Curiosity is fine, overpromising is not. A title the video does not deliver on wins the click and loses the session: viewers leave early, and YouTube uses that behaviour when deciding who to show the video to next. The honest version of the same hook usually costs a few percent of click-through rate and keeps the rest of the funnel intact.
Yes. Generation returns eight titles. Review mode assesses one existing title with an editorial rubric and alternatives. Both share the free AI-generation allowance.
Yes — they are written to be pasted, not to be examples. That said, the generator only knows the sentence you typed, so it cannot add the details that make a title yours. Dropping in the real number, product, or place you cover is usually a one-word edit and a noticeable improvement.
Shorter and blunter than long-form. A Short's title shows in search, on your channel page, and in the feed caption — but the first frame does most of the selling, so the title's job is to state the payoff in a few words rather than build curiosity. Generate with the same topic sentence and pick the option that still works cut to five or six words.
Paste your channel's public handle under the results and we compare the eight titles against the last 50 videos on that channel. If one of them repeats something you have already published, we name the old video, show its view count and link to it, and tell you which words overlap. It needs nothing but a handle anyone could look up — no sign-in, and we never read anything that is not already public. A repeat is not automatically a mistake; often the better move is to retitle the old video rather than publish a second one, which is why we point out the overlap instead of removing the option.
Renaming a published video is normal and is often the cheapest fix for one that underperformed. Give the change several days before you judge it, and change either the title or the thumbnail — not both at once — so you can tell which one moved the numbers.
