Tagsora
Open the app

Importing your library into Tagsora

Updated

On this page

Tagsora imports the file you exported from your previous app. No accounts to connect, no permissions to grant: you upload the file and that is it.

Step-by-step guides

Supported formats

Format Where it usually comes from
Bookmarks HTML Pocket, Chrome, Firefox, Edge, Safari, Raindrop
CSV with a url column Instapaper, Readwise, Raindrop, your own spreadsheets

From a CSV, Tagsora reads (when present): url, title, tags, the date (time_added, created, timestamp, date, add_date, saved…), and the highlight columns (highlight, highlights, selection, quote) with their note (note, annotation, comment) and colour.

If the CSV has no header row, Tagsora looks for the column containing addresses and uses that one.

Where import lives

In the web app: app.tagsora.com → Settings → Import. The mobile apps do not have it yet.

The file is parsed in your own browser before anything is sent: you will see how many links were detected and can cancel if the number looks wrong.

They appear immediately, but unprocessed: no summary, no automatic tags, no reader mode. Each one is processed the first time you open it.

This is not a hidden limitation, it is the decision that makes importing free. Running 5,000 links through AI at once would cost a lot of money and most of them never get opened. By deferring it:

  • Importing is instant and consumes none of your AI quota, however big the file.
  • You only pay for processing what you actually read.
  • On the free plan up to 100 new links are processed per day. If you open more, the rest stay pending and retry the next day: no bookmark is ever lost.

Duplicates

If a URL was already in your library it is not imported twice: the one you already had is kept, with its tags and notes. The final summary counts them separately as “already saved”, not as failures.

You can import the same file twice without worrying.

Highlights

Imported highlights carry the quoted text but not its position. Tagsora re-anchors them onto the article the first time you open it, by finding the quote in the text.

If the original article has changed since you highlighted it, the quote may not be found: the highlight stays saved and stays searchable, but it is not painted onto the text.

Common problems

“No links found in the file.” Almost always an unopened .zip, a CSV whose address column is not named url, or a JSON file (not a supported format — export as HTML or CSV instead).

Fewer links imported than the file contained. Rows without a web address are dropped, as are non-http(s) ones: browser-internal pages (chrome://, about:) and local files (file://).

A very large file. There is no practical limit. The import runs in batches and you can close the tab once the summary appears.

Subscriptions (OPML)

If you come from an RSS reader or a podcast app, export your subscriptions as OPML and import them in one go under Sources → Import subscriptions. The file’s folders are kept as categories and the sources you already followed are detected as duplicates. The free plan allows 30 sources: if the OPML brings more, the app tells you before starting and lets you choose which. More in Sources, podcasts and radio.

Related