TikTok Feed API
Use the TikTok Feed API to retrieve a page of public video records without a video, account or tag input. Inspect descriptions, authors, media and available engagement fields.
1 credits per non-empty successful request; empty results and failures are free. Each page is billed separately.
Each page requests 20 records; the returned number can vary. Next page is a separate billed request.
https://api.goanyapi.com/api/v1/tiktok/feed?cursor=0&count=20Run a request to see the visual result
The response is organized into metrics, lists and tables that match the JSON on the right.
Run a request to see the response here
Fill in the demo parameters to preview an example response for free. Sign in to send live requests with your own parameters.
Start with a feed page, then read each record
The TikTok Feed API provides an entry point when you want available public video records without choosing a specific video, creator or hashtag first. The response contains a page of posts, with their descriptions, authors, media and available public counts. It can supply records for a feed viewer or help you choose an individual post to inspect further. The request does not select a personal TikTok account. See the API documentation for the authentication and response reference. For a defined content category, use the Explore API.
A request without a target ID still returns individual posts
A first-page request can be written as GET /api/v1/tiktok/feed?cursor=0&count=20, with your GoAnyAPI API key supplied as a Bearer token in the Authorization header. This is an input example, not a newly measured result. The response’s data.items[] contains individual post records. Start with a record’s id and desc, identify its author, and then inspect the counts and video fields that are available. The array position is its position in this returned page; it is not a global popularity rank. If you want more information about a selected post, query its URL or ID with the Video Details API. That is a separate lookup rather than a continuation of the feed page.
- data.items[].id / data.items[].desc
- These identify and describe a returned post. They can be used to label a feed item and match it to a later detail lookup. The description is not a transcript, a topic classification, or an explanation of why the video was recommended.
- data.items[].author
- Read the author’s id, uniqueId, nickname, avatars and verification flag when returned. A handle and a display name are different fields. This author is the creator of the post, not the TikTok account whose personal recommendations were queried; no such account is selected by this request.
- data.items[].createTime
- The creation-time field describes the post. Your application’s request or save time is a separate observation and must be recorded by your own workflow. A post appearing in a newly retrieved page is not evidence that the video was published at the time you made the request.
- data.items[].stats / data.items[].statsV2
- Available playCount, diggCount, commentCount, shareCount and collectCount describe public engagement. They are not unique audience, watch completion or conversion measures. Keep the statistics objects separate, preserve zero values, and distinguish an absent count from a measured zero rather than filling every blank with the same value.
- data.items[].video
- The media object may contain covers, playback and download addresses, duration and dimensions. Use only the returned fields when building a preview. An available download address does not guarantee a permanent or watermark-free file. The adjacent field relationship illustration adds no new records and is not a TikTok screenshot.
- data.hasMore / data.nextCursor / data.pageSize
- Together these explain continuation and the configured request size. They do not provide a total number of videos or establish that another request will return unseen records. Keep the request cursor with the saved response so you can recognise repeated cursors and explain where your page came from.

A feed page has specific inputs and continuation limits
The only query parameters are cursor and count. The starting cursor is 0 and the request count is fixed at 20. There is no account, country, interest, keyword or date selection. pageSize records the request setting; use items.length for the actual returned entries, which can be fewer. Continue only if hasMore=true and nextCursor is valid and different from the cursor just used. Pass that exact returned value as cursor. Without both conditions, stop rather than adding the page length or inventing an offset. A response does not promise an endless feed or a complete video history.
Use the returned page as structured video data
Build a feed viewer
Display descriptions, author information and available covers from the returned page. Preserve missing, null and empty fields; a partial record can still contain useful information.
Inspect individual videos
Move from a feed record to video details or an author profile as needed. Those endpoints return their own data and are billed as separate requests.
Keep request costs visible
Each nonempty successful feed request is billed independently at the configured rate shown in the Tester. Empty results and failures are free. Viewing the saved example sends no new query.
Frequently asked questions
No personal account is selected by this contract, so describe the output as a returned public feed page. It does not promise to reproduce what you see after logging into TikTok, or what another person sees in their For You page.
The feed endpoint only accepts cursor and the fixed count. It has no country, interest, account, category, keyword or date parameter. For one of the defined content categories, use the separate Explore API.
No. This response has no guaranteed global popularity order or national chart. Any themes or trends you derive require your own analysis and clearly described observations. Sorting the returned counts in your application does not turn this page into a platform-wide ranking.
No. This is a synchronous GET query. A next page is a separate request using a valid new returned cursor when hasMore is true. The result reports creditsUsed and remainingCredits for that request; there is no asynchronous task to poll.
Continue with a video or author
Video Details retrieves one post, User Profile returns public account information, and User Posts lists available posts by profile URL or username.