Submit a TikTok video and get a row: the id, the author, the caption, when it went up, and the engagement counts. That is the shape. What we cannot show you is a filled-in example - every run behind this page submitted a placeholder URL rather than a real video, so the honest answer about what it returns on live content is that we have not tested it.
All six inputs were the TikTok homepage, a profile page, an incomplete path, or a placeholder id like .../@user/video/123. Every field came back empty, which tells you nothing about live content. We are not going to present that as a verdict on the service.
The four runs pointed at a homepage or a profile returned no data. The two pointed at a correctly-shaped but non-existent video id returned item doesn't exist (removed / private). Two distinct answers to two distinct mistakes is a good sign - it parsed the URL and told us which kind of wrong it was.
Runs up to 3 July returned ten columns. Runs from 14 July returned twelve, adding saves and reposts. Nothing was removed or renamed. If you built an importer against the older shape it still works, but it is now missing two columns - pin the ones you need rather than relying on position.
On all six runs url was byte-identical to query. It is not a canonical or resolved address - it is what you submitted, handed back. Do not rely on it to normalise short links or redirects.
Read from the header of a real export of this service, in sheet order. Only three of the twelve carried a value in our runs, and the reason is the inputs rather than the service - see the note below.
no data, or item doesn't exist (removed / private). Check this column before treating a blank row as a failure.saves, and absent from the earlier shape for the same reason.query - your input handed back, not a canonical URL.Twelve columns per row · CSV, JSON or Excel
Six runs sit behind this page and not one of them submitted a genuine TikTok video URL. The inputs were the homepage, a profile, an incomplete path and three placeholder ids. Every data column is therefore empty, and that is a fact about our inputs rather than about the service. We are not going to claim it is blocked, and we are not going to claim it works - we have not tested it on live content, and saying otherwise either way would be inventing a result. The 500 free rows exist precisely so you can settle this on a video you care about.
likes, comments, shares, plays, saves and reposts are separate columns rather than one blended score, so you can weight them yourself. Saves and reposts in particular say something different from likes.
Re-run the same URL list on a schedule and keep the counts against the run date. A single reading of a play count is close to meaningless; a series is where the shape of a campaign shows up.
The description column already distinguishes a missing video from a bad URL. Running an archive of links through it flags the ones that have been removed or made private, which is the cheapest form of link audit there is.
One-time, on signup, with no card. Given that we could not test this on a real video, the free tier is the sensible way to find out what it does for you.
Billed on rows actually returned, at the same rate as every other service here. No service-specific pricing.
Credits do not expire monthly, so a scheduled campaign check costs almost nothing and commits you to nothing.
The sibling service. Its metric columns did not fill and the page documents exactly why - worth reading if you are evaluating TikTok as a source.
Short-form creators usually publish on more than one platform, and YouTube gives up its numbers.
Interest over time for the topics your videos are about - the demand side of the same question.
Find the videos and creators ranking for a topic, then bring the URLs back here.
Not because the service failed - because we never asked it the right question.
Six runs, six placeholder inputs. The TikTok homepage, one profile page, one incomplete path, and three video ids that were either obviously fake or genuinely non-existent. None of them is a real video, so every data column came back empty and there is nothing to show. That is a gap in our testing, not a finding about the product, and we would rather label it correctly than dress it up as a limitation.
It matters because two of our other TikTok services genuinely are blocked - the hashtags and search scrapers both return a status explaining that TikTok serves those lists from a request-signed API. It would be easy, and wrong, to assume the same wall here. Nothing in these six runs shows that. What they do show is a service that read each URL, worked out what kind of thing it was pointed at, and answered differently for a profile than for a missing video. That is the behaviour of something that is working, being handed bad inputs.
So the honest position is: we do not know what this returns on live content, and the 500 free rows are there for you to find out on a video you actually care about. Our own terms are the same as on every other service here - publicly available pages only, nothing behind a login, no third-party trackers on the data layer, and exports auto-delete after 30 days.
query, video_id, author, description, create_time, likes, comments, shares, plays, saves, reposts and url. One row per URL you submit.saves and reposts. Nothing was removed or renamed, so an older importer still works - it is just missing two columns. Pin the columns you need rather than relying on position.no data for inputs that were not videos and item doesn't exist (removed / private) for a video id that does not exist. Different behaviour, and consistent with a service that is working.no data or item doesn't exist (removed / private). Read it before treating an empty row as a silent failure; it distinguishes a bad URL from a video that has gone.We could not test this properly, and we have said so. Your first 500 rows are free, so the answer costs you nothing to get.
The TikTok Videos Scraper takes a video URL and returns a row of twelve columns: query, video_id, author, description, create_time, likes, comments, shares, plays, saves, reposts and url. Six runs sit behind this page, and they show two things clearly. The first is that the schema grew: runs up to 3 July returned ten columns, and runs from 14 July returned twelve, adding saves and reposts. Nothing was removed or renamed, so an importer written against the older shape still loads - it is simply missing two columns, which is a good argument for pinning the fields you need rather than reading by position.
The second thing is a gap in our own testing, and we would rather label it than hide it. Not one of the six runs submitted a genuine TikTok video URL. The inputs were the TikTok homepage, a profile page, an incomplete path, and three video ids that were either obviously placeholder or genuinely non-existent. Every data column is therefore empty, and that is a fact about what we asked for rather than a finding about the service. There are no engagement figures, author names or example rows anywhere on this page, because producing any would mean inventing them.
It is worth separating this from our other TikTok services. The hashtags and search scrapers are genuinely obstructed: both return a status column explaining that TikTok serves those lists through a request-signed API rather than in the page HTML. Nothing like that appeared here. Instead the service answered differently depending on what it was pointed at - no data for a homepage or a profile, and item doesn't exist (removed / private) for a correctly-shaped video id with no video behind it. Distinguishing two failure modes is what a working parser does, so the reasonable reading is a functioning service handed bad inputs, not a wall.
One further observation for anyone building on it: the url column was byte-identical to query on every run. It echoes your input rather than resolving a canonical address, so it will not normalise short links or follow redirects for you. And description, which is meant to carry the caption, carried a status string in all six runs - read it before treating a blank row as a silent failure, because it is what tells you whether the video is gone or the URL was wrong. Publicly available pages only, no third-party trackers on the data layer, and exports auto-delete after 30 days. Your first 500 rows are free and need no credit card, which is the cheapest way to answer the question this page honestly cannot.