Best TikTok Scraping APIs in 2026: Real Cost by Depth
A TikTok data API has a sticker price and a real price, and they are usually different. A "$0.50 per 1,000" profile call becomes a very different number once you ask for the videos, comments, media and transcript that a real pipeline needs. Most comparison posts list providers, quote a per-1,000 rate, and stop. This one normalizes cost to the depth of the payload, which is the number that actually hits your invoice.
I read every competitor rate below off a live pricing page on 25 August 2026, and I present Serpent (the API this site sells) as one option among several. Where I say "I", I mean my normalization method — never a recommendation in Serpent's favour.
Short answer: the real cost of a TikTok record is a function of depth, not a single per-1,000 number. Serpent publishes that ladder explicitly: Basic $0.50 (profile and video, no include_*), Enriched $3.00 (any include_*, plus hashtag, search and music), and Media $10.00 per 1,000 (include_media=1, never free-eligible). Apify's actors quote a lower base but add comment, follower and transcript add-ons; the honest comparison is the depth you actually collect times the rate.
How I compared
I judged TikTok providers on the three things that decide a pipeline: cost per 1,000 at the depth you will actually collect, whether the author is the true creator rather than a re-post handle, and whether the API tells you when a list is incomplete. I normalized every provider to per 1,000, opened each vendor's pricing page on 25 August 2026, and I treated "free tier" and "paid plan" as separate rows because they price very differently.
I did not publish an accuracy percentage. A synthetic figure tells you nothing about the specific creator you are profiling, and it is the number I least trust from the vendors with the most to prove.
The cost-per-depth ladder
This is the table I could not find assembled anywhere else. The same 1,000 creator profiles at three depths, using each provider's published rate for that depth. "Profile only" means no videos; "profile + 30 videos" means the profile plus 30 of the creator's posts; "enriched" adds media, comments or transcript.
| Provider | Profile only /1K | Profile + 30 videos /1K | Notes on the add-ons |
|---|---|---|---|
| Serpent Basic | $0.50 | $0.50 (no include_*) | Profile and video, no included groups. |
| Serpent Enriched | $3.00 | $3.00 | Any include_*, plus hashtag, search and music. |
| Serpent Media | $10.00 | $10.00 | include_media=1; never free-eligible. |
| Apify TikTok Profile Pro | $3.00 (paid) | ~$10.00 | Profile + 30 videos ≈ $10/1K; profile + 100 videos ≈ $31/1K. |
| Apify All-in-One | $1.50 (Business) | $1.50 (Business) | 11 modes in one actor; free tier higher ($2.40 profile, $5.00 video). |
| Apify Clockworks | $1.70 (Business) | $1.70 + add-ons | Comments +$0.50/1K, follower/following +$1.50/1K, transcript $27/1K. |
| Apify TikTok Search Scraper | $0.15 | $0.15 | Keyword search results only, pay-per-result. |
Apify rates read from each actor's live pricing page 25 Aug 2026. "Profile + 30 videos" for Apify Profile Pro uses its published worked example (profile event + 20 billable videos after 10 free). Serpent tiers are its published rate card.
The point is not that one provider is cheaper. It is that a "$0.15 per 1,000" search result does not do the same job as a "$3.00 per 1,000" enriched profile record. The right comparison is depth collected times rate, and the add-on math from the provider that splits every enrichment (comments, follower lists, transcript) is the honest ceiling to budget against.
On Serpent the depth is surfaced as three named tiers rather than hidden in add-ons, so the cost of an enriched payload is visible before you build against it. A call that returns no usable data is refunded.
Author truthfulness
The author field is where TikTok scrapers quietly differ, and it is rarely explained. A video returned from a creator's page may come from the creator's own feed, a repost, a liked list, or a playlist. If a provider does not track the source of a record, the author it returns can be the account you searched rather than the person who made the video.
On Serpent the author field is the true author of the video, and the response carries the context that lets you trust it and the flags that let you act on it: music, hashtags, mentions, labels, is_ad, duet_enabled and stitch_enabled. Media and the enriched groups are only present when you ask for them with include_*, which is why the enriched tier is priced separately.
In a neutral comparison, what this means is: when you compare providers, ask whether the author is the original uploader or the page you searched, and whether the response tells you the source of the record. That single distinction decides whether your creator graph maps to real accounts.
List honesty
TikTok is a list of lists — profile groups, hashtags, searches, songs — and a list can be incomplete without a transport error. The difference between a pipeline that silently skips rows and one that knows it skipped them is whether the API exposes the list shape. On Serpent every list carries items[], result_count, total_count (null when TikTok does not state one), has_more, next_cursor, truncated, restricted and complete. When it cannot return everything you asked for, the top level adds a partialResults block with requested, returned, and a reason when the response is short by more than 40%.
The synchronous item cap is 180, and deeper runs go through the async form rather than being chopped. That is an honesty contract, not a feature: a bulk crawl built on a list you never checked for completeness is a crawl you cannot trust.
Most of the market returns a tidy array and lets you find the gap in your own analysis. Neither approach is wrong; the difference is whether the gap is visible in the response or only in your results.
The providers, fairly
- Serpent. Profile, video, hashtag, search and music. True author, explicit Basic / Enriched / Media tiers, and a list envelope that names shortfalls. Prelaunch.
- Apify — TikTok Search Scraper. Pay-per-result keyword search at $0.15 per 1,000. Cheap for flat search lists, but it search only — no profile depth.
- Apify — TikTok Profile Pro. The deepest profile-and-catalog actor: bios, contacts, reposts, liked, sponsored detection, and up to 2,000 videos per creator. Pay-per-event, so the bill scales with depth.
- Apify — All-in-One. Eleven modes in one actor, HTTP-only, and a broad output. The "from $1.50" Business rate is per mode, and the free tier is much higher per mode.
- Apify — Clockworks TikTok Scraper. The most established TikTok actor, with a large user base. The base per-result rate is low, but every enrichment — comments, follower lists, transcript — is an add-on.
- RapidAPI wrappers. Often the cheapest entry point by headline (monthly subscriptions, some from $0), but listings churn and a wrapper's maintenance and data quality vary.
- Scrapingdog TikTok. A credit-based TikTok product; the value of a call depends on the plan you buy, and it is not priced per-result the way the marketplace actors are.
Which one to pick
Decide on the depth you need before the rate.
- You want flat keyword search lists — Apify's TikTok Search Scraper at $0.15 per 1,000 is hard to beat.
- You need deep creator research with contacts and analytics — Apify Profile Pro, and budget for the video depth.
- You want a single integration across many modes — Apify All-in-One.
- You want an explicit depth ladder, a true author, and a list that tells you when it under-delivered — put Serpent, with its documented list shape, on the shortlist. It is prelaunch, and the Basic / Enriched / Media card is published so you can plan.
Before you choose any of them, pull one real hashtag or profile and read the flags the way I do: is the author the true creator, and does the list tell you what it did not return? The endpoint reference and the Python walkthrough show how to read those flags instead of trusting a clean array.
FAQ
Why does the cost of a TikTok record change so much by depth?
A profile call is cheap, but a profile plus its videos, comments and media is a much larger payload. Many providers price every enrichment as an add-on or a separate event. On Serpent the tier is explicit: Basic (profile and video, no includes) $0.50 per 1,000, Enriched (any include_*, plus hashtag, search and music) $3.00 per 1,000, and Media (include_media=1) $10.00 per 1,000.
Is the author field always the true creator of a TikTok video?
Not always. On a re-posting account the source of a video can be a repost, liked list or a playlist, and some providers fill the author from the page context rather than the original uploader. On Serpent the author is the true author of the video, and the response also carries flags like is_ad, duet_enabled and stitch_enabled, plus the media and hashtags only when you ask for them.
How do I know whether a TikTok list is complete?
Read the list envelope. Every Serpent list carries items[], result_count, total_count (null when TikTok does not state one), has_more, next_cursor, truncated, restricted and complete. A short response adds a top-level partialResults block with requested and returned, and a reason when it is short by more than 40%.
How many items can one synchronous TikTok call return?
Up to 180 items on Serpent. Deeper runs use the async form rather than silently truncating a list, so a deep crawl stays truthful about how much it actually collected.
Does the Serpent TikTok API support regions?
Pass an ISO-2 country code. The default is us, and India is normalized to us. Media is never returned without include_media=1, and it is not free-eligible.
Is the Serpent TikTok API live today?
It is in prelaunch. The endpoints and the tier card are documented now so you can plan, and the page tells you how to be notified when they ship.
