Best TikTok Scrapers: Scrape Video Data

A TikTok scraper collects public data from TikTok, including video metadata, profile details, engagement metrics, and comments, without using TikTok’s official API.

We tested four web scraping APIs on the same 1,000 TikTok post URLs, for 4,000 requests in total, and measured the success rate and completion time of each.

TikTok scraping benchmark

• Zyte had the highest success rate and the shortest time. 98.4% of its 1,000 TikTok post requests succeeded, at 18 seconds per request.
• TikTok split the four providers into two groups. Zyte and Oxylabs stayed above 96%, while Nimble (73.2%) and Decodo (71.7%) failed on more than one in four requests.
• Nimble was nearly as fast as Zyte at 19 seconds, while Oxylabs took 31 seconds and Decodo 51, almost three times Zyte’s time.

See our methodology for validation criteria and success rate definitions.

Best TikTok scrapers free trial

TikTok scrapers review in detail

Bright Data’s TikTok Scraper API provides three dedicated endpoints for collecting structured TikTok data at scale:

• Profile endpoint: Collect profile data including nickname, biography, is_verified, followers, following, videos_count, likes, and engagement metrics such as awg_engagement_rate, comment_engagement_rate, and like_engagement_rate. Supports two input methods: direct profile URL or discovery via TikTok search URL (filterable by country).
• Posts endpoint: Extract detailed post-level data including description, hashtags, play_count, share_count, collect_count, comment_count, video_duration, video_url, music, and carousel_images, alongside the creator’s profile details. Supports four input methods: direct post URL, by profile URL (with date range and post-count filtering), by keyword or hashtag, and by TikTok discover URL.
• Comments endpoint: Retrieve per-comment data including comment_text, num_likes, num_replies, comment_id, and full commenter details (commenter_user_name, commenter_id, commenter_url), tied back to the source post via post_url, post_id, and post_date_created.

On TikTok, Oxylabs came within two points of Zyte’s success rate at 96.8%, but its requests took 31 seconds on average. The setup was the Realtime endpoint of the Web Scraper API, with the universal source and HTML rendering. Oxylabs sends back the page without parsing it, and the description, author and counters are read from the markup or the embedded JSON.

Decodo’s dedicated TikTok scraper collects posts, comment threads and search results by country or keyword. Our benchmark ran on Decodo’s generic Web Scraping API with premium proxies instead.. In that setup, 71.7% of the 1,000 TikTok post requests succeeded, the lowest rate in the test, and each request took 51 seconds on average.

Zyte API had the best TikTok result in our benchmark on both measures: 98.4% of the 1,000 post requests succeeded, and the average request took 18 seconds. Requests used the browserHtml option, which loads each TikTok page in a headless browser and returns the final HTML. Post fields come from that HTML or from the JSON block TikTok embeds in the page.

Nimble answered almost as fast as Zyte, at 19 seconds per request, but only 73.2% of its 1,000 TikTok post requests succeeded, so about one in four failed. We tested its general-purpose Web API on default settings, with rendered HTML. Nimble has no TikTok-specific endpoint and relies on its residential proxy network and fingerprint handling to reach TikTok pages. 

Octoparse offers multiple pre-built TikTok scraper templates for collecting post, profile, and comment data directly from TikTok’s public pages.

Unlike API-based tools such as Bright Data or Apify, Octoparse uses visual automation that replicates real user interactions through its browser emulator. Each template supports configuration for:

• Batch input (up to 10,000 TikTok URLs)
• Custom page size (50–200 results)
• Export options (Excel, CSV, JSON, or Google Sheets)

What data can you scrape from TikTok videos?

A TikTok page can be scraped as rendered HTML, where the post data sits in the page markup or in an embedded JSON block, or as structured JSON from a dedicated TikTok scraper. Public TikTok pages contain the following data:

• Videos: description, hashtags, author, post date, duration, view, like, comment, share and save counts, music, video URL and carousel images.
• Profiles: nickname, bio, verification status, follower and following counts, video count and total likes.
• Comments: comment text, like and reply counts, commenter username and ID, and the post the comment belongs to.

Don’t miss our benchmarks and data-driven insights. The button opens Google; selecting AIMultiple confirms that you wish to see AIMultiple more often in Google search results.

Add as preferred source

TikTok scraper benchmark methodology

We tested four web scraping APIs on the same list of 1,000 public TikTok post URLs. Each API received every URL once, so the benchmark made 4,000 requests in total. We sent one URL per request and did not batch them, so each completion time belongs to a single post. We saved every response and scored the saved responses after the run. When a selector needed fixing, we could rescore without sending the requests again.

Each API was asked for the rendered page with these settings:

• Zyte: Zyte API with browserHtml, which loads the page in a headless browser and returns the final HTML
• Oxylabs: Web Scraper API through the Realtime endpoint, with the universal source and HTML rendering
• Nimble: Web API with default settings and rendered HTML
• Decodo: Web Scraping API through the generic endpoint, with headless HTML rendering and the premium proxy pool

All four APIs returned the TikTok page as HTML, but the post data is not always in the visible part of that page. TikTok often sends a mostly empty layout and stores the post details as JSON inside a script tag on the same page. A scraper that reads only the visible content sees an empty page and counts the request as a failure, even though the data is there. To avoid this, we first looked for the author, description and engagement counts in the visible HTML with CSS selectors. When the visible HTML was empty, we read the same fields from the JSON in the script tag. The view count is only available in that JSON.

A response counted as a success when we could read the author, the description or the engagement counts from it, either from the visible HTML or from the JSON. An HTTP 404 also counted as a success, because the video had been deleted and the API reported that correctly. Any other status code outside the 200 to 399 range counted as a failure. Some APIs return 200 even when TikTok itself returned an error. In those cases, we used TikTok’s status code instead of the API’s.

Completion time is the average time per request, calculated over successful responses only, so a request that fails quickly does not lower a provider’s average.

Cite this benchmark

Pick the format that matches where you're publishing. Pasting the link version into your CMS preserves the backlink.

@misc{dogan2026,
  author = {Dogan, Sedat and Şipi, Nazlı},
  title  = {{Best TikTok Scrapers: Scrape Video Data}},
  year   = {2026},
  month  = jul,
  howpublished    = {\url{https://aimultiple.com/tiktok-scraping}},
  note   = {AIMultiple. Retrieved July 24, 2026}
}

Changelog

10 updates

All10General9Methodology1

202620252024

1. Jun 26

   Added TikTok scraper benchmark results comparing success rates and metadata fields for Bright Data, Apify, and Decodo.

2. Mar 31

   Expanded the Bright Data TikTok Scraper API section with detailed fields and input methods for its three endpoints.

3. Feb 18

   Added benchmark results to the Comparison of the best TikTok scrapers section.

4. Feb 18

   Added the TikTok scrapers methodology section.

5. Feb 3

   Updated the introduction to mention TikTok Shop, device-integrity fingerprinting, and Agentic AI scraping tools.

Sedat Dogan

CTO

Follow On

Sedat is a technology and information security leader with 20 years of experience in software development, network infrastructure and cybersecurity. Sedat:
- Has 20 years of experience as a white-hat hacker and development guru, with extensive expertise in programming languages and server architectures.
- Is a board advisor at a VC investing in early-stage technology firms and at Ödeal, a regional digital payment platform serving 125,000 merchants.
- Has led the technology infrastructure and cybersecurity of seven national elections, and has been recognized in the cybersecurity Hall of Fame by global technology leaders including Twitter.

View Full Profile

Researched by

Nazlı Şipi

AI Researcher

Follow On

Nazlı is a data analyst at AIMultiple. She has prior experience in data analysis across various industries, where she worked on transforming complex datasets into actionable insights.

View Full Profile

Be the first to comment

Your email address will not be published. All fields are required. Comments are left in their original language.

Name

Email Address

Comment

0/450

Post Comment