Bring your library from Goodreads, StoryGraph or aNobii to Diavazo
If you have years of reading logged in another app, you don't have to start over. Export your library there as a CSV file, pick it in Diavazo, and the server does the rest: it looks up every book, puts it on the right shelf and tells you at the end what it couldn't find.
Where the import lives
Open your Profile, tap the Settings gear, then Your data and Import from Goodreads & co. While your library is still empty, the same page is also behind the Import button on Home and the Import a library link under the Want to read shelf.
The page lists the three sources, with two lines on how to get the file from each. Then there is one button, Pick the CSV file. You don't say which app it came from: Diavazo recognises it from the file's first row, the one with the column names. If the file doesn't look like an export from any of the three, it says so and nothing starts.
Getting the file from each app
Goodreads
The export only works on a computer, on the goodreads.com website: the app and the mobile site don't have it. Go to My Books, then Import and export, then Export Library. Goodreads prepares the file; when it's ready, download the CSV. That is the file to pick in Diavazo, as it is, without opening and re-saving it.
StoryGraph
On the StoryGraph website, your account settings have the library export (Export StoryGraph library). The result is a CSV here too: download it and pick it in Diavazo.
aNobii
On anobii.com, the CSV export of your library is in your profile settings. aNobii doesn't publish the layout of its file, so Diavazo finds the columns by name, in Italian or English: title, author, ISBN, reading status, rating, comment, start and finish dates, pages, year and publisher. If a column is missing, that detail just doesn't come across.
A useful detail: if you opened the CSV in a spreadsheet set to an Italian or other European locale and saved it again, its columns may be separated by semicolons instead of commas. That's fine, Diavazo reads it.
What comes across
From each row of the file Diavazo takes:
- the shelf: read, currently reading, to read, and on hold or did not finish where the source app tells them apart. In Diavazo they become Read, Reading, Want to read, On hold and Didn't finish;
- the dates: the finish date and, from StoryGraph and aNobii, the start date too. If you read a book more than once on StoryGraph, the latest reading counts;
- the rating, rounded to the nearest half star (StoryGraph also uses quarters);
- the review, cleaned of the HTML Goodreads puts in it;
- the page count, how many times you read it when the file says, and from Goodreads and aNobii the year and publisher, which help if the book has to be added by hand.
A book marked as read with no date stays undated. It is not counted as finished today, which would skew your year's statistics: you find it in your timeline among the books read on an unknown date, and you can add the date from the book's page.
The import is quiet. It creates no feed posts, so your followers don't see a hundred books finished at once. For how shelves and statuses work, read Shelves and statuses.
Nothing of yours is overwritten
If a book in the file is already in your Diavazo library, the import can only move it forward: want to read, then reading or on hold, then read. Never back. A book you already marked as read in Diavazo stays read; one you didn't finish stays that way.
The same goes for ratings and reviews: the file only adds them to books that don't have one yet. If you gave a book four stars in Diavazo and three on Goodreads, the four stay.
That is why you can import the same file twice, or import from Goodreads and then from StoryGraph: no duplicates appear and you lose nothing. Diavazo also remembers which book each row turned out to be, so the second time is faster.
While the import works
As soon as you pick the file, the page tells you how many books it found. Then the work moves to the server, in two steps: first it finds the books in the catalogue, then it adds them to your library. A thin bar shows how far along it is, and after the first books an estimate of the time left appears, such as About 5 minutes left. The estimate comes from that import's own pace, so at first it reads Working out how long it will take…
You don't have to stay on the page. You can leave it or close the app: the import keeps going, and its progress stays in a bar at the bottom of every page in Diavazo. When it finishes, the bar says so and you get a notification with the number of books added; tapping it opens the report.
One import runs at a time per reader. If you start another while the first is working, the new one is queued and starts on its own as soon as the other ends.
Books are looked up in Open Library, in Diavazo's own catalogue and, as a last resort, in Google Books. If a book has a valid ISBN but no catalogue knows it, Diavazo adds it to its own catalogue with the title, authors, ISBN, pages, year and publisher from the file. Your ratings, dates and reviews stay yours alone.
The report and the books not found
At the end you see three numbers: books added, updated and up to date, meaning already there with nothing to change. Below, if there are any, comes the list of books not found, with a tick box next to each and Select all at the top.
Tick the books you want to deal with and choose what to do:
- I'll add it (or I'll add them). With one book you choose between Search the catalogue, which opens the search with the title and author already typed, and Add it manually, which opens the new book form filled in with what the file had. Either way, the book you pick or create takes the row's status, dates, rating and review. With several books at once, Diavazo writes one question for ChatGPT or Claude with all the titles; you paste the answer back and check the books one by one before they go in.
- Hand to Diavazo. The book goes to the team, who add it to the catalogue before the next automatic try. The row reads handed to Diavazo, and when the book is ready you get a notification: it is already in your library, with the row's details.
- Don't add the selected. Choose Not this time and those books leave the list, but the next import asks again if they're still there; choose Never: don't ask again and every later import leaves them out, highlights too. Settings → Library lists them, and lets you ask again.
Books you leave alone stay to decide. That's fine: the next automatic try looks for them anyway.
The next-day retry
Google Books gives Diavazo a limited number of searches a day, shared by every reader. With a big library, some books may not get looked up there straight away. They are not lost: they stay with your import and the server tries them again on its own the next day, when the quota renews, without you importing again. A countdown at the top of the report says when the next try runs.
The retry shows up as a report of its own, with the line Another try, on its own, at the books your import left. You only get a notification if it found something. Leftover books are kept for up to a week, then the import is considered closed.
Starting from scratch
If an import went wrong, say you picked the wrong file, you can clear your library and run it again cleanly: Settings, Your data, Clear library, with your password. What goes and what stays is explained in Exporting your data and clearing your library. The library can't be cleared while an import is running.
If your books are on a Kobo or a Kindle instead, there is another route that also brings progress and highlights: see Kobo and Kindle.