Hi everyone! I just signed up for ResearchRabbit after seeing it mentioned in a few threads here. It looks amazing for discovering new papers, but I'm hitting a wall right at the start.
My entire library is in Zotero (hundreds of papers, all organized with tags and folders). I really don't want to manually add even a few key papers to get started in ResearchRabbit if I can avoid it. The onboarding wasn't super clear on this.
So my basic question is: can I import my Zotero library directly? If so, how does it work? Does it bring over just the citations, or notes and tags as well? If I *can't* import directly, what's the best workaround you've found?
Sorry if this is a super newbie question! I'm still figuring out my workflow for my thesis. Any guidance would be a huge help 🙏
Good question! From what I've seen, ResearchRabbit doesn't have a direct Zotero import, which is a real pain point. It's designed more as a discovery engine starting from a few seed papers.
For a workaround, you could export a collection from Zotero as a BibTeX (.bib) file and then try to import that. I'm not sure if ResearchRabbit supports .bib imports directly though - you might have to use an intermediate tool or website to convert it. It'll likely only bring over basic citation data, not your folders or notes.
If that fails, manually adding 5-10 of your most crucial "seed" papers might be the fastest way to jumpstart the recommendation engine. Once it starts, you can add more from its suggestions pretty quickly. Good luck with the thesis!
Hi, I'm in the same boat with my Zotero library. I tried the BibTeX export workaround yesterday, and it was pretty clunky. It only pulled in titles and authors, none of my tags or notes.
So I ended up just manually adding a few key papers like you mentioned. It took five minutes, and the recommendations started flowing pretty quickly after that. Have you tried that yet? I'm also curious if the tags you lose in the import matter for how ResearchRabbit suggests things.
The tags don't matter one bit to ResearchRabbit's engine, and that's your clue to the real problem here. Their algorithm works off the paper's metadata - the citation network, the abstract, the references - not your personal taxonomy. The fact that your meticulously organized tags and notes get stripped out in any import attempt isn't a bug in your process, it's a feature of their design.
You've both stumbled on the actual workflow: manually seed it with a handful of core papers. That five minute manual add is faster and more reliable than any janky export-convert-import chain that will fail on half your entries anyway.
The bigger issue is you're now maintaining two separate libraries. Your Zotero is your system of record. ResearchRabbit is just a discovery scout. You find papers there, you add the promising ones back to Zotero. Trying to force a full sync between them is a fool's errand. Treat them as separate tools.
The "five minute manual add" is the bait in the trap. It works great for your first session, when you're just playing with a toy set of seed papers. The problem surfaces when you're six months into your thesis and ResearchRabbit's discovery scouting has added 300 new, unsorted items to your "library" there, completely detached from your actual system of record in Zotero. Now you've got two diverging sources of truth and a manual sync problem that's an order of magnitude worse.
The stripped tags are a symptom, not the disease. They reveal the tool isn't built for library management, only for generating suggestions. That's fine, until you need to reconcile what it found with what you're actually keeping.
Trust but verify.