Answers

What does it cost to pull 1,000 posts from X, TikTok, Weibo or Xiaohongshu through an API?

Short answer

Through okdata keyword search, 1,000 posts cost $0.30 on X, $0.068 on TikTok, $0.20 on Weibo and $1.00 on Xiaohongshu. Each call is one page of results ($0.006 for about 20 X posts), and results already carry likes, comment counts and the author, so most jobs need no extra calls.

Say you are writing a launch report and want 1,000 posts per platform that mention the product. The bill depends on two numbers per platform: the price of one search call and how many posts that call returns.

1,000 posts, platform by platform

Platform Endpoint Per call Posts per call 1,000 posts
X X post search $0.006 20 $0.30
TikTok TikTok video search $0.002 30 $0.068
YouTube YouTube video search $0.004 20 $0.20
Weibo Weibo post search $0.002 10 $0.20
Bilibili Bilibili video search $0.002 20 $0.10
Douyin Douyin video search $0.02 20 $1.00
Xiaohongshu Xiaohongshu note search $0.02 20 $1.00

All seven platforms at 1,000 posts each come to {{cost:x.posts.search50+tiktok.videos.search34+youtube.videos.search50+weibo.posts.search100+bilibili.videos.search50+douyin.videos.search50+xiaohongshu.notes.search*50}}.

For comparison, X's own pay-per-use API lists post reads at $0.005 each on its pricing page as of November 2026, about $5 per 1,000 posts. The official API is the better choice when you also need to post, need full-archive search, or want a direct agreement with X.

One page, trimmed

curl -G "https://api.okdata.online/v1/x/posts/search" \
  -H "Authorization: Bearer $OKDATA_API_KEY" \
  --data-urlencode 'query="your product" lang:en' \
  --data-urlencode 'sort=latest' \
  --data-urlencode 'fields=id,text,published_at,metrics.likes,author.username'

Pass page.next_cursor back as cursor for the next page. Every page is billed on its own, and the amount is in meta.billing.amount.

Where the estimate goes wrong

  • Platforms only let a keyword search go so deep. If you run out before 1,000, add a second sort order or narrower keywords rather than paging further.
  • Full text and every image cost extra on some platforms: Xiaohongshu note detail is $0.02 per note. Fetch it only for the posts you will read in full.
  • Metrics a platform hides come back as null and are listed in meta.unavailable; the call is still a normal, billed result.

If the posts feed a trend report, the trend research guide shows how to group them. Open-source tools such as twscrape or MediaCrawler cost nothing per call but need accounts and upkeep, which is worth it mainly for one-off pulls by someone who enjoys the maintenance.

Updated