Examples
Open-source Python scraperMediaCrawler and similar
Open-source RedNote MCPxiaohongshu-mcp
Third-party data platformQiangua, Xinhong, Huitun
Browser extension (MediaClaw)MediaClaw
Comparison
Choosing a RedNote (Xiaohongshu) scraper: Python libraries, open-source MCP, data platforms or a browser extension
The short answer: for a platform-wide creator database and campaign history, use a data platform such as Qiangua, Xinhong or Huitun. If you code and need large scheduled crawls on a server, use an open-source scraper such as MediaCrawler. To have AI publish posts for you, use the open-source xiaohongshu-mcp. To collect the posts, comments and video transcripts you can see, then analyze them in Excel, Lark or an AI agent, use a browser extension such as MediaClaw. MediaClaw is our product, so this comparison sticks to facts each project documents publicly.
Four kinds of tools compared
Based on each project's website or repository, compiled October 2026. Prices and features change; check each official site.
| Open-source Python scraper | Open-source RedNote MCP | Third-party data platform | Browser extension (MediaClaw) | |
|---|---|---|---|---|
| Examples | MediaCrawler and similar | xiaohongshu-mcp | Qiangua, Xinhong, Huitun | MediaClaw |
| Where data comes from | A script drives a browser through pages | Its own headless browser visits pages | A database the vendor has already indexed | The pages open in your own Chrome |
| Login | QR code, cookie or phone, saved by the script | Scan a QR code in its browser; cookies stored locally | No RedNote account needed | Your everyday logged-in account; cookies never handed over |
| Setup | Python, Playwright and other dependencies; coding required | Binary or Docker, then configure an MCP client | Sign up and query on the web | Install the extension; no code |
| What you get | Raw posts, comments and profiles, across platforms | Search, post details and comments, plus publishing and likes | Creator databases, historical trends, brand and campaign analysis | Posts, comments, profiles and search results, plus video transcripts and image text |
| Output | CSV, JSON, Excel, databases | Handed to the AI in your MCP client | Reports and rankings inside the platform | Excel, CSV, Markdown, Lark, or an agent such as Codex |
| Cost | Free; MediaCrawler states learning and research use only, with a paid Pro edition | Free and open source (Apache-2.0) | Subscription; see each official site | Free extension; transcripts use credits, agent access needs a membership |
Where data comes from
Open-source Python scraperA script drives a browser through pages
Open-source RedNote MCPIts own headless browser visits pages
Third-party data platformA database the vendor has already indexed
Browser extension (MediaClaw)The pages open in your own Chrome
Login
Open-source Python scraperQR code, cookie or phone, saved by the script
Open-source RedNote MCPScan a QR code in its browser; cookies stored locally
Third-party data platformNo RedNote account needed
Browser extension (MediaClaw)Your everyday logged-in account; cookies never handed over
Setup
Open-source Python scraperPython, Playwright and other dependencies; coding required
Open-source RedNote MCPBinary or Docker, then configure an MCP client
Third-party data platformSign up and query on the web
Browser extension (MediaClaw)Install the extension; no code
What you get
Open-source Python scraperRaw posts, comments and profiles, across platforms
Open-source RedNote MCPSearch, post details and comments, plus publishing and likes
Third-party data platformCreator databases, historical trends, brand and campaign analysis
Browser extension (MediaClaw)Posts, comments, profiles and search results, plus video transcripts and image text
Output
Open-source Python scraperCSV, JSON, Excel, databases
Open-source RedNote MCPHanded to the AI in your MCP client
Third-party data platformReports and rankings inside the platform
Browser extension (MediaClaw)Excel, CSV, Markdown, Lark, or an agent such as Codex
Cost
Open-source Python scraperFree; MediaCrawler states learning and research use only, with a paid Pro edition
Open-source RedNote MCPFree and open source (Apache-2.0)
Third-party data platformSubscription; see each official site
Browser extension (MediaClaw)Free extension; transcripts use credits, agent access needs a membership
Choose by the job you need done
Most people get stuck using one kind of tool for a job another kind does better.
Find creators, check campaigns and brand data
You need a database someone already built, not your own collection. Use a data platform such as Qiangua, Xinhong or Huitun.
- Filter creators by followers and category
- See historical trends and competitor campaigns
- No account of your own visiting pages
Large, scheduled, multi-platform crawls
You can code and want jobs running on a server. Use an open-source scraper such as MediaCrawler, and read its usage terms.
- Control concurrency, proxies and storage yourself
- Covers Douyin, Bilibili, Weibo and more
- You maintain login state and anti-bot changes
Let AI publish and interact
Your goal is automated account operations. Use the open-source xiaohongshu-mcp, which publishes image and video posts, comments and likes.
- Runs from MCP clients such as Claude Code and Cursor
- One account cannot stay logged in on multiple web sessions
- RedNote only
Research specific content, then write from it
You want the posts, comments and video transcripts you can see, exported or analyzed by AI right away. Use a browser extension such as MediaClaw.
- One click on post, profile and search pages
- Timestamped video transcripts and image text
- Works with Codex, WorkBuddy and Doubao for Work
When MediaClaw is the wrong choice
Picking the wrong tool wastes more time than having none. MediaClaw is not a good fit for these needs.
| Your need | Why it doesn't fit | Better fit |
|---|---|---|
| Platform-wide creator database and campaign history | MediaClaw only collects pages you visit; there is no pre-indexed database | Qiangua, Xinhong, Huitun and similar |
| Unattended 24/7 crawling on a server | Collection runs in your own browser, which has to stay open | MediaCrawler and other open-source scrapers |
| Auto-publishing posts, likes and comments | MediaClaw collects, extracts and analyzes; it does not publish | Open-source xiaohongshu-mcp |
| Kuaishou, Bilibili, Weibo and other platforms | Only RedNote and Douyin are supported today | Multi-platform scrapers such as MediaCrawler |
Platform-wide creator database and campaign history
Why it doesn't fitMediaClaw only collects pages you visit; there is no pre-indexed database
Better fitQiangua, Xinhong, Huitun and similar
Unattended 24/7 crawling on a server
Why it doesn't fitCollection runs in your own browser, which has to stay open
Better fitMediaCrawler and other open-source scrapers
Auto-publishing posts, likes and comments
Why it doesn't fitMediaClaw collects, extracts and analyzes; it does not publish
Better fitOpen-source xiaohongshu-mcp
Kuaishou, Bilibili, Weibo and other platforms
Why it doesn't fitOnly RedNote and Douyin are supported today
Better fitMulti-platform scrapers such as MediaCrawler
FAQ
Collect a batch of the posts you want to study
Free extension, no code. Export right away or hand it to AI.