RedNote (小红书) Search Scraper — Fast Keyword Search, Filters & Full Note Metadata (2026)

Search RedNote (小红书 / Xiaohongshu) by keyword — full note content, author profile, engagement counts, image gallery, and playable video stream URLs in clean structured JSON — pick between relevance / most-popular / newest sorts, filter by note type and recency.

⚡ Built for speed. No headless browser, no per-page page-load wait, no captcha-solving overhead. 500 notes typically arrive in under 30 seconds; 1,000 notes in under a minute. Most other Xiaohongshu scrapers drive a headless Chromium instance — this one doesn't.

Copy to your AI assistant

zen-studio/rednote-search-scraper on Apify. Scrapes Xiaohongshu (RedNote / 小红书) search results by keyword. One row per note with author profile, engagement counts, image gallery, and playable video stream URLs. Call ApifyClient("TOKEN").actor("zen-studio/rednote-search-scraper").call(run_input={...}), then client.dataset(run["defaultDatasetId"]).list_items().items. Requires keywords (string[]); each keyword runs independently with its own maxResults budget. Full spec: GET https://api.apify.com/v2/acts/zen-studio~rednote-search-scraper/builds/default (Bearer TOKEN) → inputSchema, actorDefinition.storages.dataset, readme. Token: https://console.apify.com/account/integrations

How to Search RedNote by Keyword

Basic — one keyword, default 100 results

{
  "keywords": ["美食推荐"],
  "maxResults": 100
}

Multi-keyword with sort and type filter

{
  "keywords": ["护肤", "skincare routine"],
  "maxResults": 200,
  "sortType": "popularity_descending",
  "noteType": "video"
}

Newest-first with recency filter — trend monitoring

{
  "keywords": ["春季穿搭", "春天穿搭"],
  "maxResults": 300,
  "sortType": "time_descending",
  "timeFilter": "1w",
  "topUpFromOtherSorts": false
}

Input Parameters

Parameter	Type	Default	Description
keywords	string[]	required	Search terms. Chinese, English, hashtags, brand names, and emoji all work. Each keyword is searched independently with its own maxResults budget.
maxResults	integer	100	How many unique notes to collect per keyword (1–1000). Cost scales linearly with results collected. Narrow or niche keywords may yield fewer than the cap.
sortType	string	general	general (Most relevant) / popularity_descending (Most popular) / time_descending (Newest first). Controls how the platform ranks results for your search.
topUpFromOtherSorts	boolean	true	When the primary sort exhausts results before reaching the cap, automatically pull additional unique notes from the other two sort modes. Recommended ON — you get more for the same per-result price.
noteType	string	all	all / video / image / live. Restrict results to a single note format.
timeFilter	string	all	all / 1d (Past 24 hours) / 1w (Past week) / 6mo (Past 6 months). Restrict to notes published in a recent window.

Accepted keyword formats

Any term the Xiaohongshu search box accepts:

Format	Example
Chinese word or phrase	美食推荐, 护肤心得, 家常菜
English word	skincare, minimalist, coffee
Brand or person name	iPhone, lululemon, 张大奕
Hashtag	#美食, #穿搭
Emoji	🍜, ✨

Each keyword is searched independently, then results are deduplicated across sort modes so you never pay twice for the same note. A single keyword typically yields 100–300 unique notes before the platform's catalogue for that query is exhausted — use multiple related keywords for broader coverage.

Pricing — Pay Per Event (PPE)

$5.99 per 1,000 results. Each note written to the dataset counts as one result. No charge for empty keywords, retries, or pagination overhead.

Event	Per note	Per 1,000
search-result (one dataset row)	$0.00599	$5.99

Free tier

5 lifetime runs, each capped at approximately 20 results from the first keyword. No credit card required. Sufficient for evaluating the output shape and content quality before purchasing.

Advanced Usage

Trend monitoring — newest notes on a rising topic

{
  "keywords": ["防晒霜推荐 2026", "防晒测评"],
  "maxResults": 200,
  "sortType": "time_descending",
  "timeFilter": "1d"
}

Schedule this daily to track how quickly a topic gains traction. The timestamp field gives exact publish time; engagement.liked_count and engagement.collected_count reflect current totals.

Brand or competitor monitoring

{
  "keywords": ["lululemon", "始祖鸟", "北面"],
  "maxResults": 100,
  "sortType": "popularity_descending",
  "timeFilter": "1w"
}

The keyword field on each row identifies which brand surfaced the note. Cross-keyword deduplication is automatic.

Influencer discovery by niche

{
  "keywords": ["极简护肤", "敏感肌护肤"],
  "maxResults": 500,
  "sortType": "popularity_descending",
  "noteType": "video"
}

Sort by engagement.liked_count or engagement.collected_count after the run to surface the most influential creators for your niche.

Maximizing yield on a narrow keyword

{
  "keywords": ["手作皮具教程"],
  "maxResults": 1000,
  "topUpFromOtherSorts": true
}

topUpFromOtherSorts: true (the default) automatically pulls unique notes from all three sort modes when the primary sort is exhausted. For niche topics this can meaningfully increase the total yield with no extra configuration.

FAQ

How many results can I expect per keyword? Typically 100–400 unique notes before the platform's index for that query is exhausted, regardless of how high maxResults is set. For broader coverage, add multiple related keywords — cross-keyword deduplication is automatic.

What does topUpFromOtherSorts do exactly? When the primary sort mode runs out of results before reaching your maxResults cap, the actor automatically continues fetching from the other two sort modes (e.g., if you chose Most relevant, it will also pull from Most popular and Newest first). Results are deduplicated by note ID across all passes. Useful for maximizing yield per keyword at no extra per-result cost.

Do the filters combine? Yes. sortType, noteType, and timeFilter all apply together. For example, sortType=popularity_descending + noteType=video + timeFilter=1w returns the most-liked video notes from the past 7 days. Narrow combinations may reduce the result count.

How fresh is the data? Live — fetched on demand at run time. No cached or stale records.

Are the media URLs permanent? The note url (e.g., https://www.xiaohongshu.com/discovery/item/<id>?xsec_token=…) is the stable reference. Image and video CDN URLs (images[].url, video.url_720p, etc.) are time-limited and typically expire within hours to a few days. Download media immediately after a run if you need to store it long-term.

What note types does this cover? All four types the Xiaohongshu search surface returns: video notes, image/text notes (called "normal"), and live notes. Filter with the noteType input to restrict to one type. The type field on each output row tells you which category a note belongs to.

Can I search by hashtag specifically? Yes — pass "#美食" as a keyword. The search behaviour and row shape are identical to free-text search.

Can I scrape a creator's profile, their follower list, or individual note comments? Not in this actor. This actor covers keyword search results only.

What is the sortSource field? It records which sort pass produced the row: general (Most relevant), popularity_descending (Most popular), or time_descending (Newest first). Useful when topUpFromOtherSorts is on and you want to know how each note was surfaced.

What's the maximum maxResults per keyword? 1,000. In practice, the platform's index for a given query typically saturates before that ceiling; topUpFromOtherSorts helps you get closer to it.

Use Cases

• Content discovery — find the highest-engagement notes in any niche to understand what resonates with Xiaohongshu's audience.
• Trend monitoring — track newly published notes on rising topics daily or hourly using time_descending + timeFilter.
• Influencer research — identify creators publishing in your category ranked by likes and collects.
• Brand monitoring — watch what users post about your brand or competitors across multiple keywords in one run.
• Competitive analysis — compare content volume and engagement across competing brands or product lines.
• Market research — gauge category interest, note frequency, and audience sentiment on Xiaohongshu before entering a market.

Support

• Bugs: Issues tab
• Feature requests: Issues tab

Legal Compliance

Extracts publicly available search results from Xiaohongshu (RedNote / 小红书). Users must comply with Xiaohongshu's terms of service and applicable data protection regulations (GDPR, CCPA, PIPL).

---

RedNote (小红书 / Xiaohongshu) search results — full note content, author profile, engagement counts, image gallery, and video stream URLs — one row per match, in clean JSON.