xiaohongshu 小红书
Read a Xiaohongshu post with an AI agent
Turn a Xiaohongshu / RedNote link into text an AI agent can read: caption, tags, every image described, on-screen text, and a transcript for video notes. No login, no cookies.
Why this one is hard
yt-dlp's XiaoHongShuIE reads note.video.media.stream, so it only handles video notes — and image notes are the majority of Xiaohongshu. Those need the note payload parsed out of the page directly. Xiaohongshu also serves a different page depending on user agent: a mobile UA gets a 202KB app-download shell, desktop gets the real note. And a cold request for a note is rejected outright until the session has been warmed.
What comes back
Caption, tags, engagement counts, every image described with the text on it, video transcript.
Plus the metadata every digest carries: title, author, platform, posted date and the source URL. Image notes (up to 60 images) and video notes. No login needed.
Use it
Paste a link like xhslink.com/o/…, xiaohongshu.com/explore/… or rednote.com/discovery/item/… into the console, or let an agent call it directly.
MCP — Claude Code, Cursor
claude mcp add --transport http linkdigest \ https://linkdigest.dev/mcp \ --header "Authorization: Bearer YOUR_KEY"
REST
curl -X POST https://linkdigest.dev/api/v1/digest \
-H "Authorization: Bearer YOUR_KEY" \
-d '{"url": "https://..."}'10 free credits when you sign up, no card. Then $5 once for 250 credits, or $9 a month for 500 — a post is 1 credit, video is 1 a minute. Anything already digested by anyone is cached and free for everyone.