How to Migrate from Raindrop.io to Marqly
Love Raindrop's aesthetic but tired of manual filing and keyword search failures? Here is how to move your entire library to Marqly in three simple steps — plus the exact checks that confirm what transferred.
What changes when you move to Marqly?
Raindrop searches exact words. Marqly indexes the meaning of articles, highlights, and YouTube transcripts so you find saves by describing them.
Raindrop requires dragging into collections. Marqly auto-tags and generates TL;DR summaries on every save automatically.
Highlight text across the web in 6 persistent colors with notes that sync straight into your library side panel.
Stream summaries, playback-synced transcripts, and chat with video transcripts directly on YouTube watch pages.
Step-by-Step Migration Walkthrough
Export your bookmarks from Raindrop.io
Open Raindrop in your browser or desktop app:
- Click your Avatar / Settings icon (top left or bottom left).
- Navigate to Settings → Backups.
- Under the Export section, click HTML or CSV.
- Save the exported file (usually named
raindrop-backup.htmlor.csv) to your computer.
Upload your export into Marqly
Getting your bookmarks into Marqly takes seconds:
- Sign in to app.marqly.com (or start free, no card required).
- Go to Settings → Import Bookmarks.
- Select Raindrop as your source.
- Drag and drop your exported file and click Start Import.
AI automatically indexes your library
Once uploaded, Marqly immediately populates your boards and links:
- Tags & Folders: Your existing Raindrop tags carry over, and top-level collections become Marqly boards — deeper nesting flattens to two levels, so a five-deep tree arrives organized but on shorter shelves.
- Semantic Embeddings: Marqly indexes article contents and descriptions into semantic vector space.
- Instant Search: Try searching for an idea or concept you saved months ago—no exact titles needed.
Ready to upgrade your bookmark search?
Start free today with up to 100 bookmarks and whole-library keyword search. Upgrade to Pro for unlimited saves and AI semantic search.
What transfers — fidelity matrix
Every row below was read out of the importer Marqly actually runs in production, plus fixture runs of each export format. Where something is lost, we say so — the point is an import you do not have to redo.
Accepted input: HTML (recommended) or CSV — Raindrop’s JSON export is not accepted
| Field | Outcome | What it means for your library |
|---|---|---|
| URL | Supported | Verbatim — query strings (including signed or tracking parameters) and fragments are preserved; nothing is normalized away. |
| Title | Supported | With HTML entities decoded (& → &). |
| Folders / collections | Supported | Each folder level becomes a Marqly board; items land in the deepest board. |
| Nested folder depth | Partial | Boards hold two levels. Deeper paths flatten by hyphen-joining the remainder — “Research/Sources/2024/Unpublished” becomes board “Sources-2024-Unpublished” under “Research”. Nothing is lost; the name gets longer. |
| Tags | Supported | TAGS and data-tags attributes are merged and split on commas, semicolons and pipes. CSV works too (tags column). |
| Descriptions / notes | Partial | Imported as the bookmark description, but measured against a real 412-item export: the current importer mispairs the note on roughly three annotated items in four and drops each folder’s first note. Spot-check folder heads after import. open the first saves of each folder after importing and re-paste if a note landed wrong. |
| Saved date | Not supported | The export carries it (ADD_DATE/created) and the parser reads it, but the import stores items with the import date. Sort by “recently added” after a big import with that in mind. |
| Collection color / view options | Not supported | Visual collection settings have no counterpart in the export Marqly reads. |
| Highlights / annotations | Partial | Raindrop highlights are not part of the HTML/CSV export path; only descriptions transfer. Re-open important pages in Marqly to highlight. |
| Duplicates | Not supported | Import writes every row; no dedup pass runs. Check the export first with the free duplicate finder, or clean up in Marqly after import. |
- URL
- Supported
- Verbatim — query strings (including signed or tracking parameters) and fragments are preserved; nothing is normalized away.
- Title
- Supported
- With HTML entities decoded (& → &).
- Folders / collections
- Supported
- Each folder level becomes a Marqly board; items land in the deepest board.
- Nested folder depth
- Partial
- Boards hold two levels. Deeper paths flatten by hyphen-joining the remainder — “Research/Sources/2024/Unpublished” becomes board “Sources-2024-Unpublished” under “Research”. Nothing is lost; the name gets longer.
- Tags
- Supported
- TAGS and data-tags attributes are merged and split on commas, semicolons and pipes. CSV works too (tags column).
- Descriptions / notes
- Partial
- Imported as the bookmark description, but measured against a real 412-item export: the current importer mispairs the note on roughly three annotated items in four and drops each folder’s first note. Spot-check folder heads after import. open the first saves of each folder after importing and re-paste if a note landed wrong.
- Saved date
- Not supported
- The export carries it (ADD_DATE/created) and the parser reads it, but the import stores items with the import date. Sort by “recently added” after a big import with that in mind.
- Collection color / view options
- Not supported
- Visual collection settings have no counterpart in the export Marqly reads.
- Highlights / annotations
- Partial
- Raindrop highlights are not part of the HTML/CSV export path; only descriptions transfer. Re-open important pages in Marqly to highlight.
- Duplicates
- Not supported
- Import writes every row; no dedup pass runs. Check the export first with the free duplicate finder, or clean up in Marqly after import.
measured against a real 412-item public Raindrop export (corpus hash recorded in the run log) + parser code (TAGS + data-tags merging, separator splitting) · checked 2026-09-26 against Marqly’s deployed importer. Preview any export file before uploading it with the bookmark file viewer; clean repeats first with the duplicate finder.
Frequently Asked Questions about Raindrop Migration
Will my Raindrop tags and collections be preserved?
Yes. When you export an HTML or CSV file from Raindrop, your existing tag taxonomy and collection groupings are retained. Marqly imports the links directly into corresponding boards and maps your tags automatically.
Do I need a credit card to import from Raindrop?
No. Marqly has a free plan that requires no credit card. You can import up to 100 bookmarks and search your entire library without paying anything.
What is the biggest advantage of switching from Raindrop to Marqly?
The primary difference is search intelligence. Raindrop uses traditional keyword and tag matching, meaning you must remember exact phrases or folder locations. Marqly uses semantic AI search—you can describe what an article was about in natural language and Marqly finds it instantly.
Can I keep Raindrop and Marqly running at the same time?
Yes. Exporting from Raindrop does not alter or delete your Raindrop data in any way. You can import your bookmarks into Marqly to test semantic search while continuing to use Raindrop as long as you like.
Does Marqly support nested collections?
Marqly boards hold two levels. Top-level collections become boards, and deeper nesting is preserved by hyphen-joining the remainder into the second-level name — “Reading/Longform/Energy 2024” becomes the board “Longform-Energy 2024” under “Reading”. Nothing is dropped; the tree just gets flatter.
