I have nothing bug praise for Zotero. Zotero is absolutely essential to my workflow as a researcher, second only to Emacs. Without Zotero, I would be spending inordinate amounts of time keeping all my papers + associated citation information organized. Zotero just takes care of it all. I love the iOS app—I read and markup papers on my iPad and everything gets synced smoothly.
I've been a paying member for a few years now. Part of it is for the storage (PDF packrat here) but mostly because I want to support development. Please consider supporting them if they help you in your work—they're worth it. https://www.zotero.org/storage
> Zotero is open source and developed by an independent, nonprofit organization that has no financial interest in your private information. With Zotero, you always stay in control of your own data.
Refreshing as a cool breeze on a hot summer's day.
Right up until one tries to set up a self-hosted server (spoiler - you can't, at least not without 'significant effort' - they themselves say that if you ask about it).
Did you sync all data to multiple computers without a Zotero account? Zotero's documentation said WebDAV could sync attachment files. But everything else required a Zotero account.[1]
Indeed. It seems there are multiple questions in these comments on this subject, so I’ll just copy my reply from the Zotero 7 release here, since I don’t think anything has meaningfully changed since then:
WebDAV support is nice to save money, but from a privacy perspective it’s a huge bummer that the sync servers get all your citation metadata. A better self-hosting story¹ is one path to resolving this. End-to-end encryption² similar to e.g. Firefox Sync is another. Zotero has a security overview³ that shows they clearly care about good practices, but it’s still bothersome to have to trust the server when many other applications have proven E2EE works great even for non-technical users⁴.
Unfortunately from the main Zotero dev’s responses, it seems clear that they have no incentive to implement either and probably never will (look, the same comment from 2½ [now 4!] years ago⁵) without some shift in circumstances (massive increase in funding, new regulatory requirements). Even if a community member implemented the entirety of either solution, dstillman can just (rightly, tbh) claim it will increase their maintenance burden when they are trying to support paying customers.
I had a very convenient setup using linked files stored in Dropbox that worked very well for 15 years. The Zotero 6 to 7 upgrade completely broke this, and modified the database so that rollback is not possible. There was no warning that this workflow would be completely broken on upgrade.
My understanding is that that works for personal use. If you want to use a group library, not so much. Which can be considered fair, as mostly organisations which should be able to help fund zotero are the ones that need group libraries.
Is there any documentation for self-hosting that you can share or point to? Um, maybe my brain is not working today...but it sure is hell is not obvious where the instructions are for setting up a self-hosted instance. :-)
Is this intentional crippling / obfuscation, or did they just bother to do the necessary work for the server-side software robust enough to run on different HW & SW setups?
Did you sync all data to multiple computers without a Zotero account? Zotero's documentation said WebDAV could sync attachment files. But everything else required a Zotero account.[1]
No you can't. I have both a Zotero account and a WebDAV server. Since bulk of the load is the attachments, doing it this way doesn't bother me, actually.
I highly recommend everyone to use Zotero. Their original marketing as being 'for academics' is entirely wrong and it is a first-in-class bookmark/knowledge manager.
There are many software recommendations that seem sort of hype-y: Obsidian, Notion, Keybase, etc. Zotero is not that and is a daily driver for me for years. It has also replaced Calibre for me although YMMV there.
I second the recommendation for Zotero, especially if you find yourself buried under PDFs. Two things make it very useful for me. First, it organizes my PDFs and lets me search them, instead of manually searching through directories. Second, it has an OCR plugin, so I can OCR old PDFs and search the text.
I haven't used Zotero in anger for a few years and can't get to my laptop right now to verify. But I used to rely on automatic exports to a folder that I sync'd elsewhere. Never used a paid Zotero subscription, never "self-hosted" it, and had many gigs of data (including PDF attachments) working fine for years.
I used "better bibtex" (?) to ensure files were reliably renamed and moved to an appropriate folder, all automatically.
A real set-and-forget setup that ran without hitch for years.
uh oh - didn't realize sync was paid (stupidly). apparently i am at 99.6% of the free tier
now i'm interested in the answer to your question - i have my own machine running 24/7 that i would love to use. i like the software enough that maybe i'd pay/donate
My kid was talking about all the papers they had to cite in their college class. I started to suggest that they check out Zotero, but they stopped me to explain that their teachers already had them all using it.
Thank you for getting the kids started off on the right foot, professor!
I've been using Zotero as my "book" organizer. I have all my epubs, pdfs, everything there. Since version 7 I think you can read PDFs within Zotero, and I love it. I keep custom labels so I easily search for stuff. The only feature I don't use is everything about citation (funny enough). Before Zotero I had everything in file system directories, but I wanted the feeling of having one place (one app) where I could see all my books by category, by read/no-read, etc.
Having said this, I will probably wait a bit before upgrading to V8 (since I use it everyday, so I wouldn't like to face bugs and the like)
My experience with Zotero was similar - I tried adding my ebook library to it as an alternative to Calibre because I really want to sort of categorize and easily reference my books and/or get like library call number groupings which is not trivial with Calibre.
I deleted it after it only found about half of my books, which incidentially is my chief problem with Calibre.
Someday I will write an indexer with either a web search tool or an LLM interface to better find info on my books but for now I just spend too much time browsing through the files which makes me sad (but not sad enough yet to overcome the laziness)
> I deleted it after it only found about half of my books, which incidentially is my chief problem with Calibre.
Just find the citation on the web like at Open Library or somewhere, grab it, and add the book as an attachment.
I wouldn't drop it because all the stuff may not be done automatically. If you're going to read the books, you should be spending hours with them. I myself only put them into Zotero when I start reading them. I don't need to crowd it with wishful thinking. It's bloated and gets slower the more entries you add.
Post Mendeley shutdown, zotero has been an awesome replacement (while not being controlled by Elsevier). Given the amount of PDF's that I see during researchy times in my life, it's been an absolute godsend. Highly reccommend!
Zotero is the best. However, if your brain is highly tuned to use Mendeley Desktop, note that they backed down on killing it, they just won't add new features (where that leaves security updates I'm not sure).
Not OP and this is mere anecdata, but on a modest several-years-old ThinkPad, Zotero was slow when my single collection started pushing over 1,000 papers, most of which had PDFs attached. Starting up would take many seconds (half a minute?) and heavy operations such as bulk-renaming would take minutes. But for day-to-day use (adding references to my collection via a browser plugin) it was fine.
Personally, I used auto-export for all additional functionality. So, I didn't use any Word (LibreOffice) plugins that hooked into Zotero or whatever. I'd just consume a giant .bib file as and when necessary.
On modern hardware Zotero is probably fine. And it's reasonably flexible. A suggestion: export/import a big refs file (plus PDF attachments) and see if it can handle your daily workload. I suspect it will.
Love Zotero, have been using it since I started out as a researcher. I've found the PDF view to have noticeably more lag than either preview or skim, but I can live with that for entire package (and can just open the papers in those readers).
Zotero is fantastic. But I really would love to see organic bibtex support. Betterbibtex works, but it would be better that it were intrinsic feature rather than an add-on.
I discovered it a few weeks ago. It solved my problem of having a hundred arxiv tabs open. Highly recommended. I'm also looking forward to trying out the new annotations feature.
I hope they've fixed the stability issues for Zotero's Google Docs plugin when citation count gets in excess of 100+. It's a special kind of terror when the bibliography breaks and fails tracking with superscripts in the text body. The resulting necessary 'save early, save often' behavior results in accumulating hundreds of manuscript versions till submission.
I'm delighted it's now become available on ARM as I now can have it on my lab computer (Pi) and the Pocket Reform that I use day to day. I still need to come up with an optimal folder system for it though.
I‘ve been using it as a general bookmark manager (think Pinboard or Raindrop) for a while now. It‘s a bit quirky, but very powerful with all the management and annotation possibilities.
You might say it was just another excuse to curate my thousands of bookmarks and recreate a new tagging structure yet again, but… well, you wouldn‘t be wrong. :-)
Another enthusiastic Zotero user here. Current library has 13,775 items and for a low yearly price one can have multi-device sync and support the project. I'm also syncing to a server I own, for complete data ownership (just in case).
My solution: auto-export to a folder then sync using your preferred method. Use the betterbibtex plugin to rename and move all necessary files. Fiddly to set up, but reliable once it's working.
This is a sneaky edit completely different from the original post I replied to, which was about how bloated zotero was and how the op had uni stalled immediately and gone back to text files on disk.
Hi @ifh-hn , would you mind sharing what features you feel might make zotero better than text files on a disk? I'm genuinely interested to learn.
I've heard of zotero maybe a year or so ago, and was curious about it, but never took the plunge. I manage the bulk of my info/knowledge base across mostly locally-saved text files, with a few other tidbits leveraging PDFs and word process files (.odt, .docx)...and really i only use the latter for pasting in screenshots. And, then of course simply synching them across devices using syncthing.
While my approach works great for the majority of the time, i can imagine there might be some functions that some other tools might bring me which i might be missing...I suppose one thing that i lack is a graph of linkages for content that might live in different, separate files but might be related, etc. So, would you be willing to share your opinion, experience for what makes zotero better than text files on a disk? :-) Thanks!
I use it in an academic sense. I use it browser plugin to one click capture a source. Zotero mostly automatically identifies the source type, author, and other metadata. It can retrieve metadata directly from files like pdfs too. Further it has ability to retrieve metadata from ISBN, doi, and other identifiers. It can export these sources to really any citation style and also integrates with other software for this purpose. It has a plugin system for extras too.
I have nothing against text files on the disk but zotero is simply much better as a knowledge/source manager.
I really appreciate you sharing those! And, i can totally see the value in all of them. I think even just the one click capture via browser is immensely worth it, let alone those other features. Thanks again!
I have nothing bug praise for Zotero. Zotero is absolutely essential to my workflow as a researcher, second only to Emacs. Without Zotero, I would be spending inordinate amounts of time keeping all my papers + associated citation information organized. Zotero just takes care of it all. I love the iOS app—I read and markup papers on my iPad and everything gets synced smoothly.
I've been a paying member for a few years now. Part of it is for the storage (PDF packrat here) but mostly because I want to support development. Please consider supporting them if they help you in your work—they're worth it. https://www.zotero.org/storage
> Zotero is open source and developed by an independent, nonprofit organization that has no financial interest in your private information. With Zotero, you always stay in control of your own data.
Refreshing as a cool breeze on a hot summer's day.
Right up until one tries to set up a self-hosted server (spoiler - you can't, at least not without 'significant effort' - they themselves say that if you ask about it).
I haven't had an issue doing this with a standard WebDAV server and this has been true since 2007 or so.
Did you sync all data to multiple computers without a Zotero account? Zotero's documentation said WebDAV could sync attachment files. But everything else required a Zotero account.[1]
[1] https://www.zotero.org/support/sync
Indeed. It seems there are multiple questions in these comments on this subject, so I’ll just copy my reply from the Zotero 7 release here, since I don’t think anything has meaningfully changed since then:
WebDAV support is nice to save money, but from a privacy perspective it’s a huge bummer that the sync servers get all your citation metadata. A better self-hosting story¹ is one path to resolving this. End-to-end encryption² similar to e.g. Firefox Sync is another. Zotero has a security overview³ that shows they clearly care about good practices, but it’s still bothersome to have to trust the server when many other applications have proven E2EE works great even for non-technical users⁴.
Unfortunately from the main Zotero dev’s responses, it seems clear that they have no incentive to implement either and probably never will (look, the same comment from 2½ [now 4!] years ago⁵) without some shift in circumstances (massive increase in funding, new regulatory requirements). Even if a community member implemented the entirety of either solution, dstillman can just (rightly, tbh) claim it will increase their maintenance burden when they are trying to support paying customers.
1: https://github.com/zotero/dataserver/issues/105#issuecomment...
2: https://forums.zotero.org/discussion/comment/380780/#Comment...
3: https://www.zotero.org/support/security
4: https://support.apple.com/guide/security/advanced-data-prote...
5: https://news.ycombinator.com/item?id=29774935
I had a very convenient setup using linked files stored in Dropbox that worked very well for 15 years. The Zotero 6 to 7 upgrade completely broke this, and modified the database so that rollback is not possible. There was no warning that this workflow would be completely broken on upgrade.
Last Time I tried to store the Database in a Cloud Synced Folder there were like 100 warnings to click away that exactly this would happen...
I think there’s a plugin called Attanger that replicates that workflow
My understanding is that that works for personal use. If you want to use a group library, not so much. Which can be considered fair, as mostly organisations which should be able to help fund zotero are the ones that need group libraries.
Is there any documentation for self-hosting that you can share or point to? Um, maybe my brain is not working today...but it sure is hell is not obvious where the instructions are for setting up a self-hosted instance. :-)
Is this intentional crippling / obfuscation, or did they just bother to do the necessary work for the server-side software robust enough to run on different HW & SW setups?
I think it's the latter
Or try to build it yourself.
Personally I use Zotero with my own WebDAV server for sync. Works as advertised, without a hitch.
Did you sync all data to multiple computers without a Zotero account? Zotero's documentation said WebDAV could sync attachment files. But everything else required a Zotero account.[1]
[1] https://www.zotero.org/support/sync
No you can't. I have both a Zotero account and a WebDAV server. Since bulk of the load is the attachments, doing it this way doesn't bother me, actually.
I do both. I have the Zotero account for the library and webdav for the pdf files.
The question was could all data be synced without a Zotero account.
My understanding then is the answer is “probably no.”
I highly recommend everyone to use Zotero. Their original marketing as being 'for academics' is entirely wrong and it is a first-in-class bookmark/knowledge manager.
There are many software recommendations that seem sort of hype-y: Obsidian, Notion, Keybase, etc. Zotero is not that and is a daily driver for me for years. It has also replaced Calibre for me although YMMV there.
I second the recommendation for Zotero, especially if you find yourself buried under PDFs. Two things make it very useful for me. First, it organizes my PDFs and lets me search them, instead of manually searching through directories. Second, it has an OCR plugin, so I can OCR old PDFs and search the text.
Do you pay for their sync or self host? If self host, what do you use?
I haven't used Zotero in anger for a few years and can't get to my laptop right now to verify. But I used to rely on automatic exports to a folder that I sync'd elsewhere. Never used a paid Zotero subscription, never "self-hosted" it, and had many gigs of data (including PDF attachments) working fine for years.
I used "better bibtex" (?) to ensure files were reliably renamed and moved to an appropriate folder, all automatically.
A real set-and-forget setup that ran without hitch for years.
uh oh - didn't realize sync was paid (stupidly). apparently i am at 99.6% of the free tier
now i'm interested in the answer to your question - i have my own machine running 24/7 that i would love to use. i like the software enough that maybe i'd pay/donate
My kid was talking about all the papers they had to cite in their college class. I started to suggest that they check out Zotero, but they stopped me to explain that their teachers already had them all using it.
Thank you for getting the kids started off on the right foot, professor!
If you use Zotero regularly and can afford it, please consider becoming a paying member.
https://www.zotero.org/getinvolved
I've been using Zotero as my "book" organizer. I have all my epubs, pdfs, everything there. Since version 7 I think you can read PDFs within Zotero, and I love it. I keep custom labels so I easily search for stuff. The only feature I don't use is everything about citation (funny enough). Before Zotero I had everything in file system directories, but I wanted the feeling of having one place (one app) where I could see all my books by category, by read/no-read, etc.
Having said this, I will probably wait a bit before upgrading to V8 (since I use it everyday, so I wouldn't like to face bugs and the like)
My experience with Zotero was similar - I tried adding my ebook library to it as an alternative to Calibre because I really want to sort of categorize and easily reference my books and/or get like library call number groupings which is not trivial with Calibre.
I deleted it after it only found about half of my books, which incidentially is my chief problem with Calibre.
Someday I will write an indexer with either a web search tool or an LLM interface to better find info on my books but for now I just spend too much time browsing through the files which makes me sad (but not sad enough yet to overcome the laziness)
> I deleted it after it only found about half of my books, which incidentially is my chief problem with Calibre.
Just find the citation on the web like at Open Library or somewhere, grab it, and add the book as an attachment.
I wouldn't drop it because all the stuff may not be done automatically. If you're going to read the books, you should be spending hours with them. I myself only put them into Zotero when I start reading them. I don't need to crowd it with wishful thinking. It's bloated and gets slower the more entries you add.
Post Mendeley shutdown, zotero has been an awesome replacement (while not being controlled by Elsevier). Given the amount of PDF's that I see during researchy times in my life, it's been an absolute godsend. Highly reccommend!
Zotero is the best. However, if your brain is highly tuned to use Mendeley Desktop, note that they backed down on killing it, they just won't add new features (where that leaves security updates I'm not sure).
https://blog.mendeley.com/2025/07/09/mendeley-is-not-going-a...
Is it easily self hostable yet?
I'm glad part of their stack is open source but I just wish they made it as easy as a compose file to run this on prem.
Zotero is very good nowadays. Its unbearably slow though, and it doesn't seem this will ever get better.
Could you share how/when it is slow? We’re considering using this at work and I’d love your feedback.
Not OP and this is mere anecdata, but on a modest several-years-old ThinkPad, Zotero was slow when my single collection started pushing over 1,000 papers, most of which had PDFs attached. Starting up would take many seconds (half a minute?) and heavy operations such as bulk-renaming would take minutes. But for day-to-day use (adding references to my collection via a browser plugin) it was fine.
Personally, I used auto-export for all additional functionality. So, I didn't use any Word (LibreOffice) plugins that hooked into Zotero or whatever. I'd just consume a giant .bib file as and when necessary.
On modern hardware Zotero is probably fine. And it's reasonably flexible. A suggestion: export/import a big refs file (plus PDF attachments) and see if it can handle your daily workload. I suspect it will.
Love Zotero, have been using it since I started out as a researcher. I've found the PDF view to have noticeably more lag than either preview or skim, but I can live with that for entire package (and can just open the papers in those readers).
Zotero is fantastic. But I really would love to see organic bibtex support. Betterbibtex works, but it would be better that it were intrinsic feature rather than an add-on.
I discovered it a few weeks ago. It solved my problem of having a hundred arxiv tabs open. Highly recommended. I'm also looking forward to trying out the new annotations feature.
Zotero had been “not good enough” for many years. Since version 7, it has become nearly perfect for me. I can’t recommend it highly enough.
I hope they've fixed the stability issues for Zotero's Google Docs plugin when citation count gets in excess of 100+. It's a special kind of terror when the bibliography breaks and fails tracking with superscripts in the text body. The resulting necessary 'save early, save often' behavior results in accumulating hundreds of manuscript versions till submission.
For those using Zotero for knowledge management or bookmarking, can you give us some examples of how?
I'm delighted it's now become available on ARM as I now can have it on my lab computer (Pi) and the Pocket Reform that I use day to day. I still need to come up with an optimal folder system for it though.
Honestly the piece of software in my life that I have the least issues with ever. So so good.
Doesn't title have space for more context like "tool to collect, organize, annotate, cite research" or something?
What a blast from the past (for me at least). When I used this 5 or so years ago (17 years ago) it was extremely effective.
Coincidentally, I was just looking at using Zotero on Linux this morning after having used it previously on Mac for some classes.
I‘ve been using it as a general bookmark manager (think Pinboard or Raindrop) for a while now. It‘s a bit quirky, but very powerful with all the management and annotation possibilities.
You might say it was just another excuse to curate my thousands of bookmarks and recreate a new tagging structure yet again, but… well, you wouldn‘t be wrong. :-)
Another enthusiastic Zotero user here. Current library has 13,775 items and for a low yearly price one can have multi-device sync and support the project. I'm also syncing to a server I own, for complete data ownership (just in case).
How do you sync to your own server? Do you mean just "some" data or did you figure out how to self host their open source (but undocumented) backend?
My solution: auto-export to a folder then sync using your preferred method. Use the betterbibtex plugin to rename and move all necessary files. Fiddly to set up, but reliable once it's working.
I think it's just WebDAV, used to be at least.
amazing tool thank you
> amazing tool thank you
This is a sneaky edit completely different from the original post I replied to, which was about how bloated zotero was and how the op had uni stalled immediately and gone back to text files on disk.
Zotero is built on top of Firefox ESR.
Im unaware of another open source tool that provides all zotero does that's lighter weight. I can guarantee it's better than using text files on disk.
Hi @ifh-hn , would you mind sharing what features you feel might make zotero better than text files on a disk? I'm genuinely interested to learn.
I've heard of zotero maybe a year or so ago, and was curious about it, but never took the plunge. I manage the bulk of my info/knowledge base across mostly locally-saved text files, with a few other tidbits leveraging PDFs and word process files (.odt, .docx)...and really i only use the latter for pasting in screenshots. And, then of course simply synching them across devices using syncthing.
While my approach works great for the majority of the time, i can imagine there might be some functions that some other tools might bring me which i might be missing...I suppose one thing that i lack is a graph of linkages for content that might live in different, separate files but might be related, etc. So, would you be willing to share your opinion, experience for what makes zotero better than text files on a disk? :-) Thanks!
I use it in an academic sense. I use it browser plugin to one click capture a source. Zotero mostly automatically identifies the source type, author, and other metadata. It can retrieve metadata directly from files like pdfs too. Further it has ability to retrieve metadata from ISBN, doi, and other identifiers. It can export these sources to really any citation style and also integrates with other software for this purpose. It has a plugin system for extras too.
I have nothing against text files on the disk but zotero is simply much better as a knowledge/source manager.
I really appreciate you sharing those! And, i can totally see the value in all of them. I think even just the one click capture via browser is immensely worth it, let alone those other features. Thanks again!