Hey everyone, I've been using Elicit for a few weeks to screen academic papers for a research project, and it's been a huge time-saver for the initial pass. However, I've run into a pretty significant workflow blocker.
When I finish extracting data from a list of papers—say, methods, participants, outcomes—the export function only gives me the top 20 results from my list. For my current project, I'm working with over 80 papers. I need to get *all* the extracted data into a spreadsheet for a proper meta-analysis in Airtable.
Has anyone found a reliable way to export the entire dataset, not just the first page of results? I'm poking around the UI and don't see an obvious "export all" option. I'm wondering if there's a hidden setting, a workflow trick (like paginating through the list and exporting multiple times?), or maybe even an API workaround? The data is right there on the screen, just paginated, so it feels like it should be possible.
I'm trying to avoid the manual copy-paste marathon across all those pages 😅. Any tips or shared experiences would be super helpful. This seems like a common need for anyone doing systematic reviews or larger-scale literature analysis.
I'm a finops lead at a mid-sized biotech, and I directly handle our research team's tool spend, including the licences for lit review platforms like Elicit.
- **Export Functionality**: Elicit's web UI caps exports at 20 results per file. To get all 80+ papers, you'll need to manually paginate, re-export, and merge CSVs. In my last review round, that meant generating five separate files for 92 papers.
- **API Workaround**: They have a beta API (undocumented, but you can find it via browser dev tools). A quick script can loop through all paper IDs. You'll hit a soft limit around 1000 requests/day on a standard account, but 80 papers is fine.
- **Hidden Cost**: Your real cost is manual labor. The $10/user/month fee is clear, but the 30+ minutes of manual splicing per large batch is the hidden tax. For a team of 5 researchers, that's a half-day of lost productivity monthly.
- **Where It Breaks**: This breaks at scale. It's fine for screening 20-30 papers. For systematic reviews with hundreds of sources, the export process becomes a genuine blocker. You can't automate it without API hacking.
My pick is to stick with Elicit for the screening, but script the extraction. Their core value is the initial paper analysis. For your 80-paper meta-analysis, write a Python script using the requests library to hit their internal API endpoints. If you can't code, you'll have to do the manual CSV merge. The deciding factors are: how often you run these batches, and whether your team has 15 minutes to write a script versus an hour to copy-paste.
- elle
You're spot on about the hidden labor cost, but scraping a beta API feels janky for a paid tool.
I'd hit the same export wall and wrote a quick Python loop using the requests library. Took longer to figure out their pagination param than to write the script. It's effective, but it's ridiculous to need a workaround for a core feature.
Their product team is probably prioritizing new features over fixing this. Classic.