Weibo data.
With the exact counts.
One key returns Weibo posts with exact repost, comment and like counts, comment threads, profiles, search back through time and the live hot-search list, all as typed JSON.
Trusted by developers shipping real products.
13,085,280 served today
Request
$ curl -X POST \ https://api.fetchlayer.dev/weibo/search-posts \ -H "Authorization: Bearer ss-your-key" \ -H "Content-Type: application/json" \ -d '{"query":"深圳","since":"2026-10-01"}'
Response
{
"posts": [{
"id": "5354374692867529",
"text": "深圳11月1日起，全市禁止机动车鸣喇叭…",
"createdAt": "2026-10-15T12:04:29.000Z",
"postedFrom": "广东",
"author": { "screenName": "1024专员", "verified": true },
"stats": { "reposts": 33, "comments": 234,
"likes": 400 },
"roundedStats": [],
"media": [{ "kind": "image", "downloadUrl": "…/weibo/media?url=…" }]
}, …49 more],
"newestFirst": true, "hasMore": true, "pagesFetched": 1
} Built for feedback workflows
Fetch the public signal. Use it wherever your team works.
Exact counts, not rounded ones
A post’s reposts, comments, likes and reads, and an account’s followers, come back as whole numbers — 23800264, not 2380万 — so week-to-week movement is something you can measure.
Search that reads back through time
Every matching post newest first, 50 to a page, with a cursor that keeps going back for years, and since and until to pin the period. These are the posts Weibo’s public search shows.
The hot-search list, with its heat scores
Weibo’s 热搜 as it stands right now: rank, topic, the exact heat score, category and when each topic joined the list, with paid placements left out unless you want them.
Where each post came from
Every post and comment carries the province or country Weibo says it was posted from, so a conversation splits by region without guesswork.
Comment threads, replies included
Most-liked or newest first, each comment with its likes and reply count, and every reply under it on request. Where a very large thread stops, the response says so.
Photos and video that open anywhere
Weibo’s own media links open only on Weibo. Every photo and video here carries a download link that opens anywhere, with Range support for seeking.
Find any account
By id, link, custom address or display name, with its bio, verification, location, join date and the totals it has received across every post.
Any link, resolved for free
Hand /resolve-url a Weibo address and it tells you whether it is a post, an account, a search or a topic, plus the ids it encodes and the route that reads it.
Read what China is saying—in English, then export it.
Prefer not to write code? Ask FetchLayer Research Chat what is trending on Weibo or what people say about a brand, and get the posts summarised in English with a clean table to download.
- ✓Ask questions in plain English
- ✓Analyze themes, sentiment, and pain points
- ✓Download the structured results as CSV or JSON
No code or setup required · Uses your FetchLayer credits
Search Weibo for posts about 华为 from the last 7 days. Summarise the main themes in English, list the 20 most-liked posts with their exact like, comment and repost counts, and export as CSV.
I read the last week of matching posts newest first. The summary groups them into five themes, and the table lists the 20 most-liked posts with the author, the province each was posted from, and exact like, comment and repost counts. It is ready to download.
weibo-huawei-7d.csv
20 posts · 8 columns · ready to download
Solutions
Build the workflow around the feedback.
The short-video side
TikTok accounts, videos, hashtags and comments, for when a campaign runs in China and abroad.
Explore →The same conversation, in the West
Posts, replies and profiles from X, in the same typed shape, for side-by-side brand monitoring.
Explore →Where the long discussion lands
Threads and comment trees about a product or a topic, for the English-speaking half of the picture.
Explore →From raw feedback to action
One request is the start of the workflow.
Store the structured response, run it on a schedule, pass it to an LLM, or feed it into the system your team already uses. FetchLayer handles retrieval so you can focus on the outcome.
- 01
Start from what is hot, or from a term
Read the hot-search list, or search a brand, a keyword or a #topic# between two dates.
- 02
Read the posts
Up to 1,000 posts in one request, newest first, each with exact counts, the author, topics, photos and video.
- 03
Go deeper where it matters
Pull a post in full, read its comments and replies, or profile the accounts behind it. Store what you read: counts are as at the time of the request.
One command.
Your agent does the rest.
Use the Weibo endpoint directly from your MCP-compatible agent. Ask it to retrieve public feedback, identify patterns, and turn the result into a useful brief without writing integration glue.
1 Run one command, then sign in
$ npx fetchlayer
◇ Agents
│ Claude Code and Codex
◇ Signed in as you@company.com
◇ Connected
│ ✓ Claude Code ✓ Codex
It opens a sign-in page in your browser. New here? Signing in creates your free account.
2 Ask
›What is trending on Weibo right now, and what are people saying about the top topic?
fetchlayer.weibo_trending
The top topic is 深圳全市全天禁鸣喇叭5年 — Shenzhen banning car horns citywide for five years — with a heat score of 1,095,555, on the list since 14:55. The most-liked comments ask what a horn is for if it can never be used. One request, one credit.
No key to copy and nothing to install. New accounts are free, with no credit card.
Use cases
What teams build with Weibo data.
Same API, endless applications. Here is what teams ship with Weibo posts, comments, profiles and the hot list.
China brand monitoring
Search your brand on Weibo, newest first, and page back as far as you need. Every post arrives with exact repost, comment and like counts, so you see what is spreading and not just what was said.
Hot-list tracking
Read Weibo’s hot-search list with the exact heat score behind every topic and store it on a schedule. Weibo shows the list as it is right now, so a history exists only if you keep one.
KOL vetting
Judge an account on its exact follower count and the reposts, comments and likes it has received across every post, rather than on the rounded figure in a media kit.
Comment mining
Read the most-liked or newest comments on a post that took off, with where each commenter posted from, and find the questions and complaints Chinese buyers repeat.
Campaign measurement
Follow a #topic# or a set of posts after launch and report reposts, comments, likes and reads per post, all exact.
Market research datasets
Search a term between two dates and build a corpus of posts with their province of posting, topics and engagement counts, for the period you actually care about.
Compare approaches
Spend time on the insight, not the collection layer.
| What you need | FetchLayer | Official API | Build it yourself |
|---|---|---|---|
| Public data with no app review | ✓ | Application required | You maintain it |
| Exact post and follower counts | ✓ | Varies by endpoint | You parse them out |
| One record shape on every route | ✓ | Varies by endpoint | You design it |
| Partial results explained, not silent | ✓ | n/a | You detect it |
| Ready for AI and MCP workflows | ✓ | Your integration | Your integration |
FetchLayer provides access to publicly available data. It is not affiliated with, endorsed by, or sponsored by the platform.
Start free. Scale when ready.
Try it free, buy credits as you go, or take a monthly plan when your volume is steady.
Free
$0
30 requests
Enough to call the platforms you care about and read real responses before you pay anything.
- Every platform and endpoint
- REST and MCP
- No credit card
Pay as you go
$1.99
per 1,000
Buy credits when you need them. They stay on your account until you use them.
- Credits never expire
- No monthly commitment
- Same access as subscribers
- ∞ req/min, ∞ API keys
Monthly plan
Starter$25
per month
15,000 requests a month, about 500 a day.
$1.67 per 1,000, $5 less than credits each month
- ∞ req/min, ∞ API keys
1 page fetched is 1 credit
A normal call reads one page. Ask for more pages and each one is one more credit.
The same on every platform
No multiplier for harder targets, and no charge per result returned.
Nothing back, nothing billed
If a request does not return usable data, it costs 0 credits.
Not sure which volume fits?
15 minutes with the founder. Tell us the workload, we tell you what it needs.
Guaranteed throughput on the platform you run on
A dedicated platform from $400/mo, plus $500 for each ~21 req/s. Every other platform on the same key. 99.8% uptime, 24-hour support on Slack or Telegram.
API reference
Every parameter, documented.
All requests are POST with a JSON body and Bearer auth header.
POST /weibo/search-posts Posts matching a keyword or a #topic#, newest first, 50 to a page, going back with a cursor. since and until limit the period; ranked tabs give Weibo’s top and hot picks.
query* string A keyword, or #topic# for a topic tab string 'latest' (default) | 'all' | 'original' | 'verified' | 'top' | 'hot' | 'videos' | 'media' | 'images' | 'combined' since / until string A date (China time) or an ISO date-time limit / cursor number | string Up to 1000 posts, or resume from a previous nextCursor. POST /weibo/trending Weibo’s live hot-search list (热搜): rank, topic, the exact heat score, category, Weibo’s new and hot badges and when each topic joined the list. Paid placements are left out unless you ask for them.
includePromoted boolean Include paid placements, marked isPromoted limit number Up to 60 topics POST /weibo/post-detail One post in full: the complete text, author, when and where it was posted, exact repost, comment, like and read counts, topics, photos and video, and the post it reposts.
post* string A post link, numeric id or short id POST /weibo/post-comments Comments on a post, most-liked or newest first, with likes, reply counts, the few replies Weibo shows inline and the province each commenter posted from. /comment-replies reads a whole reply thread.
post* string A post link, numeric id or short id sort string 'hot' (default) | 'latest' limit / cursor number | string Up to 1000 comments, 20 to a credit. POST /weibo/user-profile An account’s profile by id, link, custom address or display name: bio, verification, location, join date, the exact follower count and the reposts, comments and likes it has received across every post.
user* string A numeric id, profile link, custom address or display name POST /weibo/user-posts An account’s posts, newest first, 50 a page, with a cursor back through roughly its newest 2,000, each with exact counts. With mediaOnly, only its photo and video posts, reaching further back.
user* string A numeric id, profile link, custom address or display name mediaOnly boolean Only posts with photos or video, going further back POST /weibo/search-users Accounts matching a name or keyword, with verification, bio, post count and the follower count as Weibo shows it in search.
query* string A name or keyword POST /weibo/resolve-url What a Weibo link points to — post, account, search or topic — with its ids and the route that reads it. Free.
url* string Any Weibo link Fields marked * are required. Full API reference →
FAQ
Common questions.
What is a Weibo API?
Here it means one HTTP endpoint that returns public Weibo (微博) data as clean JSON: one post in full, the comments on it and the replies under a comment, an account’s profile and newest posts, post search by keyword or #topic#, account search and the live hot-search list. You send a link, an id, a name or a term with a Bearer token, and get back typed records with text, timestamps, the province a post was made from, exact engagement counts, authors, topics, photos and video. There is no Weibo developer application to file, no review and no Weibo account. Everything returned is public on Weibo.
Do I need a Weibo account or developer approval?
No. You send your FetchLayer key and nothing else. FetchLayer is not affiliated with, endorsed by, or sponsored by Weibo, and the Weibo name is used only to identify the platform the data comes from.
Are the counts exact or rounded?
Exact. A post’s reposts, comments, likes and reads come back as whole numbers, and a profile’s follower count is the exact figure — 23800264, with Weibo’s rounded 2380万 beside it. roundedStats names any counter that came back rounded; normally it is empty. The one place Weibo shows only a rounded follower count is account search, and the response says so: user-profile gives the exact figure.
Can I search Weibo by date?
Yes. since and until take a date, read as that day in China time, or an ISO date-time. On the default tab, latest, and on all, original and verified, search returns every matching post newest first, 50 to a page, and nextCursor keeps going back — for years if the term is old enough — up to 1000 posts in one request. top, hot and videos are Weibo’s ranked picks, one page per time range.
Does search return every post?
It returns the posts Weibo’s public search shows, which can leave out posts Weibo filters from search. That is usually most of what is said about a brand or a topic, but it is not a census, and we do not present it as one. Search has no region filter.
Can I read an account’s full history?
Back through roughly its newest 2,000 posts: user-posts returns them newest first, every kind of post, 50 a page, with a cursor for older ones. With mediaOnly it lists only the posts that carry photos or video, which reach further back. To read what an account said about a subject over a period, search for the subject with since and until.
How complete are a post’s comments?
Comments page 20 at a time, most-liked or newest first, up to 1000 in one request, and comment-replies reads every reply under a comment. A very large thread stops at a few hundred comments — about 300 most-liked first, or about 740 newest first — however many Weibo counts, and notes says so, so a partial read never passes for a whole one.
How does billing work?
By results returned, and every response states it as pagesFetched. A post, a profile, an account search and the hot-search list are 1 credit each. Comments and replies are 1 credit per 20. Search is 1 credit per 50 posts returned, minimum 1, and the combined tab is 9. An account’s posts are 1 credit per 50 returned, or with mediaOnly 1 per 48. Parsing a link with /resolve-url is free, and each photo or video download is 1 credit.
Can I download the photos and videos?
Yes. Weibo’s own media links open only from Weibo’s pages, so every photo and video carries a downloadUrl that opens anywhere, with Range support so a player can seek. Each request to it is 1 credit.
Can my AI agent use it?
Yes. All nine routes are MCP tools on the same FetchLayer MCP server: weibo_post_detail, weibo_post_comments, weibo_comment_replies, weibo_user_profile, weibo_user_posts, weibo_search_posts, weibo_search_users, weibo_trending and weibo_resolve_url. They use the same key and are billed exactly like REST.
What is not available?
Reposts lists and follower lists. An account’s posts beyond roughly its newest 2,000. A region filter on search. And historical snapshots: counts and the hot list are as they stand when you ask, so tracking movement means storing what you get back.
Start pulling Weibo data today.
Get a free API key, make your first request, and put public feedback to work in your own product or research workflow.
Used by developers at
Individual developers or teams — not official partnerships