Literature Radar: Track New Papers and Explore Related Research Inside Zotero (Explore + Track)
If you research anything in 2026, you already know the problem: thousands of new papers land on arXiv, PubMed, bioRxiv and medRxiv every single day, and the pace is only accelerating as AI tools speed up how fast science gets written. Your alert system for all of this is probably still a Google Scholar email from 2019, or a PubMed saved-search you stopped opening months ago.
Meanwhile, the place you actually work — reading, annotating, and writing — is Zotero. That mismatch is why we built Literature Radar, a new two-part literature monitoring engine inside the PapersGPT plugin for Zotero:
- 🔍 Literature Exploration (Explore) — instant, AI-assisted discovery across seven research databases, starting from a topic you type or the paper currently open in your reader.
- 📡 Literature Track (Track) — subscribe to the topics you care about and receive daily or weekly digests of new papers, delivered inside Zotero with a red-dot inbox.
Both are available today. Here is what they do, how they differ, and how they compare with the alert tools you're probably using now.
What Is Literature Radar?
Literature Radar is a researcher-facing monitoring engine built into PapersGPT for Zotero. In one sentence:
Literature Radar watches arXiv, Semantic Scholar, PubMed, Europe PMC, Crossref, bioRxiv and medRxiv for papers matching the topics you choose, and brings results — either on demand (Explore) or on a schedule (Track) — directly into your Zotero workspace.
Two design decisions make it different from a web service:
- Your queries go directly from your machine to the source APIs. There is no monitoring server in the middle collecting what you search for. Radar calls arXiv, Semantic Scholar and PubMed straight from your Zotero client.
- Failures are visible, never silent. Each source reports its own status (rate-limited, timeout, error). If PubMed is slow today, Radar tells you instead of pretending there were no results.
🔍 Literature Exploration: From One Paper to the Whole Field
Explore answers the question every researcher asks weekly: "what should I read next?"
- Start anywhere. Type a topic ("CRISPR base editing off-target prediction"), or launch Explore with no query at all and Radar uses the paper currently open in your reader as a seed.
- Two engines at once. Explore runs keyword search across the databases and queries Semantic Scholar's recommendation engine with your seed paper — so you get both "everything matching these words" and "work that is actually related to this paper."
- Search in your own language. Non-English queries are automatically translated into optimized English keywords before hitting arXiv and PubMed, because those APIs only match English text well. 中文检索、日文検索都能直接用。
- Biomedical-aware routing. Queries that look biomedical automatically route through PubMed, Europe PMC, bioRxiv and medRxiv; CS/physics queries lean on arXiv and Semantic Scholar. You don't configure any of this.
- Every result is actionable — title, venue, year, source and link, one click to open or save into your library.
📡 Literature Track: Your Topics, Watched Every Day
Explore is for the moment. Track is for the months that follow — the slow, unglamorous work of staying current that nobody has time for.
- Subscribe to a topic. Anything you can search, you can track: a method ("transformer interpretability"), a disease, a rival lab's output, a compound.
- Pick a cadence. Each subscription runs Daily or Weekly. A lightweight scheduler inside Zotero checks your topics and collects new matches.
- Read the digest, not the firehose. New papers for each topic land in a digest view inside Zotero with a red-dot notification — title, source, and link for every new match. Mark read when you've caught up.
- No server, no lock-in. Scheduling runs locally in your Zotero process. When Zotero is closed, nothing runs — the next launch catches up on due subscriptions. Your research interests never leave your machine.
Explore vs. Track: When to Use Which?
| 🔍 Explore | 📡 Track | |
|---|---|---|
| Question it answers | "What's related to this / what exists on X right now?" | "What's new on my topics since yesterday/last week?" |
| Runs | On demand, instant | On a Daily/Weekly schedule |
| Input | A typed topic or the paper you're reading | A saved topic subscription |
| Best for | Literature reviews, related-work sections, checking novelty of your idea | Long-running projects, thesis watch, competitive labs, staying current in a moving field |
| Think of it as | A research search engine | A personal alerting agent |
They're designed to be used together: Explore to map the field, Track to keep the map updated.
How Does Literature Radar Compare to Google Scholar Alerts and PubMed Emails?
| Literature Radar | Google Scholar Alerts | PubMed Saved-Search Emails | |
|---|---|---|---|
| Lives where you work | ✅ Inside Zotero | ❌ Email inbox | ❌ Email inbox |
| Multi-database | ✅ 7 sources, auto-routed | ❌ Google index only | ❌ PubMed only |
| Related-work recommendations from a seed paper | ✅ Yes | ⚠️ "Cited by" chains only | ❌ No |
| Non-English queries | ✅ Auto-translated | ⚠️ Mixed results | ❌ English only |
| Digest cadence control | ✅ Daily or Weekly, per topic | ❌ As-it-happens emails | ⚠️ Fixed options |
| Query privacy | ✅ Client-direct, no middle server | ☁️ Google account | ☁️ NCBI account |
| Works from your current PDF | ✅ Yes | ❌ No | ❌ No |
Scholar and PubMed alerts are fine for one or two high-signal queries. The moment your list grows past that, email becomes a graveyard — which is exactly the gap a Zotero-native monitor fills.
Getting Started with Literature Radar
- Install PapersGPT in Zotero (Zotero 7 or later): download the plugin here.
- Open the Radar panel — the 🔍 Explore and 📡 Track New controls sit in the PapersGPT chat panel.
- Type a topic and hit Explore. When the results look right, subscribe to the same topic with Track New and pick Daily or Weekly.
Free tier: every user gets 5 Explore runs + 5 Track runs to test the workflow with no payment and no API key. Explore is included in every paid plan; Track subscriptions are a Premium / Business feature (track unlimited topics with Daily or Weekly digests).
FAQ
Does Literature Radar run when Zotero is closed?
No. Track subscriptions are checked by a scheduler that runs inside your Zotero process — nothing executes on a server while you're away. When you reopen Zotero, due subscriptions run once and your digest is ready. This is a deliberate privacy choice: your topics are checked from your machine, directly against the source APIs, and never stored on ours.
Which databases does Literature Radar search?
Seven: arXiv, Semantic Scholar, PubMed, Europe PMC, Crossref, bioRxiv and medRxiv. Queries are routed automatically — biomedical topics hit the biomedical databases, preprint-heavy fields hit arXiv, and Semantic Scholar's recommendation engine is used for seed-paper related work.
Can I search in Chinese or another language?
Yes. Non-English queries are automatically translated into optimized English keywords before querying, since arXiv and PubMed match English text far better. Results still come back with full metadata and links.
Is Literature Radar free?
Every user gets 5 free Literature Exploration runs and 5 free Literature Track runs — enough to test the full workflow. Explore is included in all paid plans; Track subscriptions (unlimited topics, Daily/Weekly digests) require a Premium or Business plan.
How is this different from just checking Google Scholar?
Google Scholar is a search box; it doesn't watch anything for you unless you set up email alerts, which arrive in your inbox as raw link lists without cadence control per topic, without multi-database coverage, and without any connection to the paper you're currently reading. Literature Radar runs where your library already lives, covers seven databases, and turns the papers you read into seeds for discovery.
Does my library data get uploaded anywhere?
No. Radar queries are made directly from your Zotero client to arXiv, Semantic Scholar, PubMed and the other APIs. Your PDFs, library contents and topic subscriptions stay on your machine.
Stop re-searching. Start watching. Download PapersGPT and point Literature Radar at your first topic today — or see the full feature list first.