Douyin Video Search Scraper (抖音搜索)

Search Douyin (抖音) by keyword and get the videos it finds, without a Douyin account, login or API key. Type one or more keywords, in Chinese or English; the run reads Douyin's video results page after page, adds the videos of its general results that the video tab did not show, removes the duplicates, and delivers each video with its link, caption, hashtags, publication date, likes, comments, shares and saves, its author (with follower count and certification) and its sound. Export as JSON, CSV or Excel, or read it through the Apify API.

$5 per 1,000 videos, plus a small start fee per run. Videos removed by your filters, duplicates and failed searches are not charged.

How deep it goes

Measured on Apify on 17 October 2026, Max results per keyword 500:

Keyword	Videos delivered
咖啡	162
露营	234
iPhone 17	168
travel china	323
健身	201
5 keywords in one run, 85 s	1,088, no duplicate

With Search related keywords on, plus filters (published in the last 180 days, 500 likes or more, videos only), 2 keywords gave 233 videos in 3 minutes: 1,424 more were found and removed by the filters, not charged.

Douyin's web search shows a limited number of videos per keyword, then says there are no more. To go further, turn on Search related keywords: when a keyword runs out, the run also searches the related searches Douyin suggests for it (up to 5, three at a time), and each video says which search found it (searchedAs). The related searches are Douyin's own suggestions and can drift from your keyword (for fashion: fashion舞蹈, but also a song title). Several keywords are searched at the same time.

Quick start

1. Type one or more keywords in Keywords, one per line (咖啡, 露营, iPhone 17).
2. Set Max results per keyword (100 by default).
3. Click Start and read the videos in the run's Output tab.

Like any Apify Actor, it can also be run and read through the Apify API, scheduled, and connected to webhooks, Make, Zapier, n8n or Apify's MCP server for AI agents.

Inputs

• Keywords: one per line, in Chinese or English. A Douyin search link (https://www.douyin.com/search/咖啡) works too. The same keyword entered more than once is searched once.
• Max results per keyword: in Douyin's order, the most relevant first.
• Search related keywords (off by default): see above.
• Published after: only videos published after a date (2026-09-01) or within a period (30 days).
• Min likes: only videos with at least this many likes.
• Content type: Douyin search mixes videos and photo posts (图文). Keep both, or only one of them.
• Max results: total videos the run delivers, all keywords together, counted exactly.

Filters are applied to what Douyin's search returns: a narrow filter on a keyword with few results can leave few videos. Videos removed by a filter are counted in the run summary and never charged.

Example, through the API:

{
  "keywords": ["咖啡", "露营装备", "travel china"],
  "maxResultsPerKeyword": 200,
  "searchRelatedKeywords": true,
  "publishedAfter": "90 days",
  "minLikes": 1000,
  "maxItems": 600
}

Output

One record per video. A real one, from 17 October 2026:

{
  "type": "video",
  "status": "ok",
  "keyword": "露营",
  "position": 4,
  "videoId": "7510585934795115803",
  "url": "https://www.douyin.com/video/7510585934795115803",
  "title": "☔️在雨中使用新的车顶帐篷露营",
  "caption": "☔️在雨中使用新的车顶帐篷露营 #户外露营#露营漫时光#携宠出游#周末去哪玩#住进风景里",
  "hashtags": [
    "户外露营",
    "露营漫时光",
    "携宠出游",
    "周末去哪玩",
    "住进风景里"
  ],
  "categories": [
    "随拍",
    "生活记录",
    "日常vlog"
  ],
  "publishedAt": "2025-05-31T12:32:42Z",
  "durationSeconds": 1035.5,
  "isPhotoPost": false,
  "imageCount": null,
  "likeCount": 4898,
  "commentCount": 483,
  "shareCount": 1580,
  "saveCount": 1208,
  "coverUrl": "https://p5-ex-gddgtc-sign.douyinpic.com/tos-cn-p-0015/o4tAakFOeIDYr6CAZcUgfgEMo6IJ95hADLwv9B~tplv-dy-360p.jpeg?…",
  "width": 1440,
  "height": 1080,
  "isAd": false,
  "sellsProducts": false,
  "author": {
    "id": "3393515805216940",
    "secUid": "MS4wLjABAAAAgt7nK28gRiIBkHUVzhCfc_LmkttZHBAtBnvaYZWt86JS8SJBVF3jckCSUU_nkyb_",
    "username": "49944574028",
    "fullName": "露营🏕女孩",
    "followerCount": 12166,
    "isVerified": false,
    "verification": null,
    "url": "https://www.douyin.com/user/MS4wLjABAAAAgt7nK28gRiIBkHUVzhCfc_LmkttZHBAtBnvaYZWt86JS8SJBVF3jckCSUU_nkyb_",
    "avatarUrl": "https://p3.douyinpic.com/aweme/100x100/aweme-avatar/tos-cn-i-0813_oQvgnEAGD5AAADpBBIQIaijnm7qzEZkPQq2Ai.jpeg?…"
  },
  "sound": {
    "id": "7510587183888206618",
    "title": "@露营🏕女孩创作的原声",
    "author": "露营🏕女孩",
    "isOriginal": true
  },
  "searchedAs": "露营",
  "scrapedAt": "2026-10-17T06:51:36.328Z"
}

Fields that need a word of explanation:

• position is the video's rank for that keyword in this run (1 = first shown by Douyin).
• searchedAs is the search that found the video: the keyword itself, or a related keyword when Search related keywords is on.
• categories are the topics Douyin files the video under, broadest first.
• isPhotoPost: a photo post (图文) rather than a video; imageCount is its number of images, durationSeconds is empty for it.
• saveCount: how often the video was saved to favorites.
• isAd: a paid placement in the results. sellsProducts: the video links to products in Douyin's shop.
• author.secUid is the stable ID of the account in Douyin links; author.username is the Douyin ID the user chose, and can change.
• author.followerCount is empty (null) for the videos found through Douyin's general results: that page does not carry the author's followers, and the run never writes a made-up 0.
• author.isVerified is true when Douyin shows a certification badge on the author; author.verification is the label it shows (a job, a brand, a shop, a media).
• coverUrl is a signed image link that Douyin keeps valid for about two weeks: download the covers you need soon after the run.
• A search that fails gets an error record, free: error says why, retryable says if a later run may work.

The run summary (videos per keyword, related searches Douyin suggested, filtered and duplicate counts, why the run stopped) is the OUTPUT record of the run's key-value store.

Go further with the other Douyin scrapers

Search is where a Douyin project starts. Pass the url of the videos you found to:

• Douyin Transcript Scraper: what is said in each video, as text.
• Douyin Comments Scraper: the comments of each video.
• Douyin User Search Scraper: accounts instead of videos, by keyword, with their full public profile.

Pricing

Pay per event:

Event	Price
Video	$5 per 1,000
Run start	$0.005 per GB of memory (1 GB minimum; $0.005 at the default 512 MB)

Never charged: videos removed by your filters, a video found by two keywords (delivered once), error records. Set Maximum cost per run in the run options and the run stops before going over it.

Limits

• Douyin's web search has no sort or date options for visitors: the order is Douyin's relevance order, and the date and likes filters are applied by the run on what the search returns.
• Video playback links are not delivered: Douyin's links from search expire within minutes. Use the url of the video.
• View counts are not delivered: Douyin's search gives 0 views for every video, so any number would be false.
• Douyin decides how many videos a keyword shows; two runs a few hours apart can return a somewhat different list.

FAQ

Do I need a Douyin account or to log in? No.

Does it work with TikTok? No, Douyin only (the Chinese app, douyin.com).

Can I search in English? Yes. Douyin is a Chinese platform, so a Chinese keyword usually finds more videos.

Why fewer videos than my limit? Douyin's search ran out for that keyword; the run says so in its status and in the run summary. Turn on Search related keywords, or add more precise keywords.

What if the run is restarted by the platform (migration or reboot)? It resumes where it was: videos already in the dataset are not delivered or charged again.

What if Douyin changes its pages? The run fails with an error that says so, instead of reporting "no videos", and nothing is charged but the start.

Is it legal? It collects public videos shown to any visitor of Douyin's search. Check the rules that apply to your use of the data, especially for personal data.

Does the Actor send anything to Quarra? Only when a run fails or runs out of time: the Actor, its build, the error type, the line of our code where it happened, and the memory and time the run was given, so we can fix it fast. Never your input, your results, your account or the run ID.