RedNote (Xiaohongshu) Scraper — Trending, Users, Hashtags & Search

The canonical RedNote / Xiaohongshu (小红书) scraper for Apify. Pull trending notes, search by keyword, scrape hashtag/topic pages, or extract a specific user's posts — with structured JSON output ready to drop into a dashboard, CRM or quant model.

RedNote (小红书 / Little Red Book) is China's dominant lifestyle super-app — "Instagram + Pinterest + Yelp + Amazon reviews" rolled into one. 300M+ MAU, ~70% female, average age 25, with an unusually high concentration in tier-1 / tier-2 mainland cities (Shanghai, Beijing, Shenzhen, Guangzhou, Chengdu). After the 2025 TikTok ban scare, an additional ~700K self-described "TikTok refugees" joined RedNote in 72 hours, accelerating its international footprint.

Crucially, every viral RedNote note is, in effect, a product review or a buying guide. Trending RedNote content is the closest publicly observable signal to Chinese consumer demand — earlier and cleaner than Tmall / JD sales rankings, and orders of magnitude more granular than Nielsen-style panel data. This actor extracts that signal as clean JSON, on demand, with no manual scrolling and no Mandarin-speaking analyst in Shanghai required.

This actor was the #1 most-requested scraper on the Apify Ideas page (45 community upvotes, 2.5× the next-highest request).

---

What you get per record

Every RedNote post returned by the actor is structured as:

Field	Type	Description
post_id	string	RedNote note id (the /explore/{id} hex token)
author	string	Note author's display name
author_id	string	Author's user id (joinable to user_posts mode)
title	string	Note title
body	string	Note caption / description text
image_urls	list	Image URLs attached to the note (up to 9)
hashtags	list	Hashtags extracted from title + body
likes_count	int	Likes (decoded from 12.3万 / 1.2k shorthand)
comments_count	int	Comment count
shares_count	int	Share count
collected_count	int	Save / collect count
engagement_total	int	Sum of likes + collected + comments + shares
posted_at	string	ISO-8601 timestamp when the note was published
url	string	Canonical xiaohongshu.com/explore/{id} deep link
category	string	beauty / fashion / food / travel / wellness / tech / finance / lifestyle
mode	string	trending / keyword_search / hashtag / user_posts
query	string	The input that produced this record (keyword / hashtag / user)
data_source	string	Provenance (httpx vs playwright, mode)
as_of_timestamp	string	UTC ISO-8601 capture time

If — and only if — every probe strategy is fully blocked by RedNote's anti-bot wall, the actor emits a single status: "maintenance" record explaining the situation and listing the sister actors you can fall back on. Your downstream pipeline never crashes. No charge is applied for maintenance stubs.

---

Four scrape modes

Set mode to one of the following:

trending

Pulls the RedNote homepage discovery feed (/explore) — the same surface a logged-out RedNote user lands on. Great for "what's hot on RedNote right now" daily monitoring. No other inputs needed.

keyword_search

Searches RedNote for one or more keywords (keywords array). Chinese characters give native-quality hits (e.g. ["口红", "护肤", "穿搭"]); English fall-backs work too (e.g. ["lipstick", "skincare", "outfit"]). Use this for brand mention monitoring, competitive product research, or category sweeps.

hashtag

Pulls notes tagged with one or more hashtags / topics (hashtags array). Provide tag text without the leading # (e.g. ["秋冬穿搭", "咖啡探店"]). Falls back to keyword search if the topic page is unavailable. Use this for tracking a specific campaign tag, a trending topic, or a community hashtag.

user_posts

Pulls recent posts from one or more specific users (users array). Accepts full profile URL (https://www.xiaohongshu.com/user/profile/{id}), a path (/user/profile/{id}), or just the bare user id. Use this for KOL / influencer monitoring, creator-economy research, or pulling a portfolio of brand-owned accounts.

---

How the actor handles RedNote's anti-bot wall

RedNote is hosted in mainland China and applies aggressive bot detection on datacenter IPs (HTTP 403/461, empty feeds, captcha walls). The actor uses a multi-strategy waterfall:

1. httpx + realistic browser headers with zh-CN locale and a real Chrome User-Agent — extracts window.__INITIAL_STATE__ when RedNote renders server-side hydration data.
2. DOM-level fallback parser — pulls /explore/{id} anchors and inferred titles when the hydrated state blob is absent or partial.
3. Playwright headless Chromium with stealth countermeasures — navigator.webdriver patch, zh-CN locale, Asia/Shanghai timezone, realistic viewport, plugin / languages spoofing. Handles JS-rendered hydration when HTML alone isn't enough.
4. Maintenance-stub fallback — graceful single-row response so consumer pipelines don't break (and no charge applied).

All paths run behind Apify's RESIDENTIAL proxy pool by default, which is strongly recommended for RedNote.

---

How buyers actually use this actor

• Beauty / fashion e-commerce ops — daily 5-keyword keyword_search monitor against the brand portfolio + 3 closest competitors. Plug into Slack via webhook for "your brand just appeared on RedNote trending" alerts.
• China market-entry consultants — weekly 200-record keyword_search sweep across 美妆 + 健身 + 生活方式 to identify whitespace categories before a client launch.
• KOL / influencer agencies — daily user_posts pull of the top 50 KOLs in a vertical to track posting cadence, engagement decay and brand-collab disclosures.
• Hedge fund consumer-discretionary desks — trending + hashtag records as a real-time alt-data feed for theses on Estée Lauder, L'Oréal, Anta, Li Ning, Pop Mart.
• Brand-safety / compliance teams — flag is_branded_partnership=true posts that mention competitors for review.
• TikTok refugees research — hashtag mode against the migration tags (#TikTokRefugee, #Xiaohongshu) to size and segment the cross-platform user influx.

---

How this compares to incumbent China consumer-trend tools

Capability	RedNote Scraper (NexGenData)	Bloomberg China Consumer Pulse	Nielsen China	Kantar Worldpanel	Hootsuite China
Per-record API access	Yes (pay-per-event)	Terminal subscription only	Annual contract	Annual contract	Seat-based SaaS
Latency	Live (≤2 min)	Daily	Monthly panel	Weekly	Daily-ish
Cost to first record	Cents	$24k+/year/seat	$80k+/year minimum	$50k+/year minimum	$9k+/year
Four scrape modes (trending / search / hashtag / user)	Yes	No	No	No	Limited
Self-serve, no contract	Yes	No	No	No	No
Pipeline-friendly JSON	Yes	CSV / terminal	PDF / Tableau	Excel	CSV

The incumbents are useful for board-level narratives; this actor is what you wire into your daily marketing-ops dashboard.

---

Pair it with the NexGenData China & Social-Listening fleet

This actor is one slice of NexGenData's purpose-built Chinese-data fleet on Apify. The other essentials:

• Weibo Hot Search Tracker — real-time Weibo trending topics ("China's Twitter").
• Bilibili Video Search — China's YouTube / Twitch — video discovery & metadata.
• China Trends Tracker — cross-platform trending aggregator across Weibo / Douyin / Baidu / WeChat.
• Eastmoney China Stock Screener — A-share fundamental / momentum data for joining consumer trends to public-company tickers.
• HKEX IPO Tracker — Hong Kong IPO pipeline for cross-border brand listings.
• Finance MCP Server — one-call MCP wrapper bundling the Chinese finance fleet for LLM agents.

Build a cross-border consumer-trend pipeline by pairing RedNote Scraper (demand signal) with Eastmoney (public-company tickers) and HKEX IPO (listing pipeline) — daily.

---

Get 20% off your first 3 months on Apify

If you're new to Apify, use this referral link to sign up and get 20% off for 3 months, then 30% ongoing. Pairs cleanly with any of the China / social-listening actors above.

---

Input parameters

• mode — trending (default) / keyword_search / hashtag / user_posts.
• keywords — array of Chinese or English search terms (used in keyword_search mode).
• hashtags — array of hashtags / topic slugs without leading # (used in hashtag mode).
• users — array of full URL / path / bare user id (used in user_posts mode).
• limit — max records to return (1-500; default 30).
• proxyConfiguration — Apify proxy (defaults to RESIDENTIAL — strongly recommended).

---

Disclaimer

This actor pulls publicly visible RedNote content using the same browser endpoints any logged-out RedNote.com visitor reaches. It does not bypass paywalls, login walls, age gates or private profiles. Buyers are responsible for compliance with RedNote's Terms of Service and applicable Chinese / EU / US data regulations in their jurisdiction.