VK Keyword Hit Search

What does VK Keyword Hit Search do?

VK Keyword Hit Search reads the public VK news feed for the keywords and dates you provide, then keeps comments on those posts that contain the same keywords. It stores one dataset row per hit and a styled Excel workbook. It does not log in as a VK user, does not read private profiles, and does not search every comment on VK. Comments are loaded only from posts that already matched.

The Actor calls the official VK API methods newsfeed.search and wall.getComments. A service access token with permission to use those methods is required. A VK ID user token limited to personal info cannot search the news feed.

Why use VK Keyword Hit Search?

VK has no single endpoint that searches comments across the whole site. This Actor does the two-step collection in one run:

• Search posts in a required time range, newest first.
• Open comments only on posts that matched, and keep a comment only when its text matches too.
• Treat OR as separate searches whose posts are merged by link.
• Treat AND, commas, Chinese commas, semicolons, and spaces as “every term must appear”.
• Stop a page loop when VK returns a short page or no next page, instead of requesting empty pages.
• Always write the Excel file, including a header-only workbook when nothing matched.
• Leave the run private until you choose to publish it. Scheduling, the Apify API, and dataset exports stay available on a private Actor.

What data can VK Keyword Hit Search extract?

Field	Type	Description
日期	text	Publish time in Asia/Shanghai
类型	text	帖子, 转发, or 评论
内容	text	Post or comment text returned by the API, with HTML tags removed
链接	link	https://vk.com/wall{owner}_{post} or the same URL with ?reply=
作者	text	User or community name
作者主页	link	Public VK profile or community page
粉丝数	number	Followers for a user, or members for a community
点赞 / 评论数 / 转发 / 互动总数	number	Counts from the API. For a comment, 评论数 is the reply count
平台	text	Always VK

The API text is the body. The Actor does not open each post in a browser to fetch another copy.

How to scrape VK posts and comments

1. Open the Actor and set the secret environment variable SKILL_VK_ACCESS_TOKEN to a VK service token. Do not put the token in the input.
2. Optional: set SKILL_VK_PROXY when this machine cannot reach api.vk.com directly.
3. Open Input. Enter keywords, a start date, and an end date.
4. Leave post pages and comment pages at 10 unless you want a shorter sample. The Actor stops early when a page is empty or shorter than requested.
5. Start the run. Open Output for the Excel file results.xlsx, or download the dataset as JSON, CSV, or Excel.
6. Read the OUTPUT record for the match mode, row counts, suggested file name, and warnings.

How much will it cost to scrape VK?

This Actor is private and has no Store price. A run spends Apify compute units for the time the container is alive, plus your own VK API quota. VK documents newsfeed.search as usable with a user token or a service key and caps a single search at the first 1000 results. Comment collection is the larger part: each matched post can request up to 10 pages of 100 comments. Lower pages and comment_pages when you only need a sample. Publishing later can use pay per usage or pay per event; the platform fee on paid Actors is separate from VK, which does not charge for these read methods.

Input

See the Input tab for every field. query, start_time, and end_time are required.

самокат AND segway keeps an item only when both words appear. camokat OR segway searches each word and unions the posts. A comma is AND, not OR. Dates without a time are Asia/Shanghai: the start date begins at 00:00:00 and the end date finishes at 23:59:59. If you omit the VK time arguments yourself, VK would silently search only the last 24 hours; this Actor refuses to run without both dates.

The token is not an input. Add SKILL_VK_ACCESS_TOKEN under the Actor's environment variables and mark it Secret.

Output

You can download the dataset in JSON, HTML, CSV, or Excel. The styled workbook is the key-value record results.xlsx. A summary JSON record is stored as OUTPUT.

{
  "日期": "2026-10-16 22:10:53",
  "类型": "帖子",
  "内容": "public post text",
  "链接": "https://vk.com/wall-170492840_23573",
  "作者": "Community name",
  "作者主页": "https://vk.com/club170492840",
  "粉丝数": 1000,
  "点赞": 3,
  "评论数": 1,
  "转发": 0,
  "互动总数": 4,
  "平台": "VK"
}

Tips

Use one page of posts and one page of comments for the first run, then raise both toward the default of 10. OR multiplies post searches, so a long date range with several OR groups takes longer. A warning that comments exceeded the page cap means that post has more comments than the run scanned; matches outside those pages are not included.

FAQ, disclaimers, and support

Which VK token works?

newsfeed.search accepts a user access token or a service key. Tokens issued for personal info only are rejected by VK. Create the token in the VK app settings and store it as a Secret. The Actor never writes it into the dataset or the Excel file.

Does a comma mean OR?

No. A comma, Chinese comma, semicolon, or space means every term must appear. Use the word OR with spaces around it when you want separate searches and a union.

Is the text the full post body?

Yes. The text fields from newsfeed.search and wall.getComments are stored as 内容. There is no second request for a different article body.

Our Actors are ethical and do not extract any private user data, such as email addresses, gender, or location. They only extract what the user has chosen to share publicly. We therefore believe that our Actors, when used for ethical purposes by Apify users, are safe. However, you should be aware that your results could contain personal data. Personal data is protected by the GDPR in the European Union and by other regulations around the world. You should not scrape personal data unless you have a legitimate reason to do so. If you're unsure whether your reason is legitimate, consult your lawyers.

Questions about a run can go to the Actor's Issues tab. Programmatic runs use the API tab on this Actor.