(I) Research API
To make it easier to independently research our platform and bring transparency to TikTok content, we built a Research API that provides researchers in the US, EEA, UK and Switzerland, with access to public data on accounts and content, including comments, captions, subtitles, number of comments, shares, likes, followers and following lists, and favourites that a video receives on our platform. More information is available
here. We carefully consider feedback from researchers who have used the API and continue to make improvements such as additional data fields, streamlining the application process, and enabling collaboration through Lab Access, which allows up to 10 researchers to work together on a shared research project.
(II) Virtual Compute Environment (VCE)
The VCE allows qualifying non-academic not-for-profit researchers in the EU to access and analyse TikTok's public data, while ensuring robust security and privacy protections. Public data can be accessed and analysed in 2 stages:
- Test Stage: Query the data using TikTok's query software development kit (SDK). The VCE will return random sample data based on your query, limited to 5,000 records per day.
- Execution Stage: Submit a script to execute against all public data. TikTok provides a powerful search capability that allows data to be paginated in increments of up to 100,000 records. TikTok will review the results file to make sure the output is aggregated.
(III) Commercial Content API
As required under the DSA, and to enhance transparency on advertisements presented on our platform, we have built a
commercial content API that includes ads, ad and advertiser metadata, and targeting information. Researchers and professionals are required to create a TikTok for Developers account and submit an application to access the Commercial Content API which we review to help prevent malicious actors from misusing this data.
(IV) Commercial Content Library
The Commercial Content Library is a publicly searchable database with information about paid ads and ad metadata, such as the advertising creative, dates the ad ran, main parameters used for targeting (e.g. age, gender), number of people who were served the ad, and more. It also includes information about content that's commercial in nature and tagged with either a paid partnership label or promotional label, such as content that promotes a brand, product or service, but is not a paid ad.