"Best TikTok API" is the wrong search. TikTok doesn't have one kind of data worth pulling — it has at least five, and the endpoint that's great for one is often the wrong shape for another. An agent built to spot trending hashtags needs something completely different from one built to price-check TikTok Shop listings.
Below is what actually separates a good TikTok data API from a bad one, and which real endpoint covers each of the five jobs.
If you're an agent (or building one) reading this rather than a human, the full machine-readable schema for every endpoint below lives at mindcase.co/skills.md.
What should you actually evaluate in a TikTok data API?
Four things matter more than a feature checklist:
- Coverage shape. Does one endpoint try to do everything, or are posts, profiles, and shop listings separate calls with separate schemas? A single overloaded endpoint usually means shallow fields on all of them.
- Login requirement. Does pulling this data need an authenticated TikTok session, or does it run against what's publicly visible? A session-based approach means account bans are your problem to manage.
- Freshness. Is a request answered live against TikTok right now, or served from a cache that might be hours or days old? For anything trend-related, a stale cache defeats the entire point.
- Pricing shape. Per-row pricing scales down naturally for a small test run. A flat per-request or subscription price doesn't — you pay the same whether you pull 10 rows or 10,000.
Which TikTok data are you actually trying to get?
These are five different jobs, not five levels of the same thing:
- Posts — videos and photo posts by profile, hashtag, search term, or sound, with engagement metrics attached to each one.
- Profiles — finding accounts by keyword, with follower counts, engagement history, and business/seller flags.
- Shop — product listings on TikTok Shop, with pricing and sold counts.
- Comments — what people are actually saying under a specific video.
- Followers — the follower or following list of one specific account.
TikTok Posts API is the one most people reach for first, and it's also the most commonly misused — pulling posts when what you actually need is the account behind them, or the comment thread under one specific video.
curl -X POST "https://api.mindcase.co/v1/data/tiktok/comments/run?wait=true" \
-H "Authorization: Bearer $MINDCASE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"params": {
"postURLs": "https://www.tiktok.com/@username/video/1234567890",
"commentsPerPost": 100
}
}'
# $0.0005 per comment returnedTikTok Comments API returns 12 fields per comment — text, likes, reply count, the commenter's handle and profile URL, and whether the creator liked or pinned it. That last pair is a cheap way to find the responses a creator actually cared about, without reading every comment yourself.
How does pricing actually work across TikTok endpoints?
Every TikTok endpoint here is priced per row, but the per-row price varies by how much work each row represents — a comment is cheap because it's a small object; a Shop product listing costs more because it carries pricing, variants, and stock data:
| Endpoint | Price | Fields per row |
|---|---|---|
| TikTok Comments | $0.0005 / comment | 12 |
| TikTok Followers | $0.001 / profile | 15 |
| TikTok Posts | $0.002 / post | 27 |
| TikTok Profiles | $0.0025 / profile | 19 |
| TikTok Shop | $0.015 / product | 15 |
There's no subscription, no minimum commitment, and no per-seat fee layered on top — pulling 50 comments to sanity-check a video costs $0.025, not the price of a monthly plan you'd only use once.
Does it require a TikTok account or login?
No, across all five endpoints. Every one of them runs against public post, profile, and shop data — no login, no app installation, no session cookie to keep alive. That matters more than it sounds: a scraper built on a logged-in session is one TikTok layout change away from breaking, and the account it's logged in as is now a liability if it gets flagged.
What can't you get from Mindcase?
Two cases.
Certified historical archive access. These endpoints read what's publicly visible on TikTok right now — the same thing a person scrolling the app sees. TikTok's own Research API is a separate, application-gated product built specifically for compliance-grade historical access. If your use case requires that certification, this isn't a substitute for it.
Data on a private account. Followers, posts, and profile details all depend on the account being public. A private account returns nothing, by TikTok's own design — no vendor gets around that, and none honestly should.
Which endpoint should you use for which job?
| Job | Endpoint |
|---|---|
| Track a hashtag or trend over time | TikTok Posts |
| Find accounts by keyword or niche | TikTok Profiles |
| Check product listings and pricing | TikTok Shop |
| Read what people are saying under a video | TikTok Comments |
| Pull one account's follower or following list | TikTok Followers |
FAQ
No. All five endpoints run against public data — no login, no session, no app install required.
No — they're five separate endpoints with separate schemas, because each one covers a genuinely different object (a post, a comment, a product listing). Combining them means calling each endpoint you need.
Live on every run. Nothing here is served from a stale cache — each request reflects what's on TikTok at the moment you call it.
TikTok Comments, at $0.0005 per comment. TikTok Shop is the most expensive per row, at $0.015 per product, since each row carries pricing, variants, and stock data.