A Connected Papers alternative makes sense once you move from "show me this field" to "find every paper I missed and prove it". Citation Map in AI for Research does that second job: backward and forward snowballing rounds across a set of papers, ranked candidates, and RIS or CSV export, for anyone writing a thesis or review.
What Connected Papers is best at
Connected Papers builds a graph of papers similar to an origin paper, using co-citation and bibliographic coupling rather than direct citation lines, from the Semantic Scholar corpus [1], [3]. Paste one DOI or title and you get a map of a few dozen papers where similar work sits close together, so sub-topics show up as clusters. Its Prior works list points to foundational papers the graph cites most, and Derivative works points to newer papers that cite many graph papers, often recent reviews [3]. For the first week on a new topic, nothing gets you an overview faster. It also finds papers that belong to the same conversation even when neither cites the other, which a citation-only tool can miss.
If that is your job, use Connected Papers. The rest of this page is for when you need something it was not designed to do.
Where AI for Research differs
Citation Map answers a different question: not "what is similar to this paper?" but "what does my set of papers cite, what cites it, and which of those did I miss?"
- Snowballing across a set, not one graph per seed. Add papers to a working set, then run a Backward, Forward or Both round. Each round reads the references or citing papers of the whole set and ranks candidates by how many set papers link to them ("Linked to 3 of your set").
- A record of what you did. Every paper carries the round it was added in (R0, R1, R2), and the map is saved to your History so you can reopen it.
- Reviews that cite your set. One button finds reviews and meta-analyses citing your papers, often the fastest check for gaps.
- Export for reporting. RIS and CSV, with the round in the CSV, ready for a reference manager or a systematic review.
- The next step in the same place. Save the set to your Library, ask questions of the PDFs, or attach the papers to a Deep Research question.
Connected Papers and Citation Map compared (checked November 2026)
Built only from each product's own pages and our code. Prices change; check the vendor's page before you buy.
| Connected Papers | Citation Map (AI for Research) | |
|---|---|---|
| Best at | Instant overview of a field from one paper | Reportable backward and forward snowballing from a set |
| How links are built | Similarity from co-citation and bibliographic coupling [1] | Direct references and citing papers; candidates ranked by links into your set |
| Data | Semantic Scholar corpus [1], [3] | OpenAlex, Semantic Scholar as fallback |
| Start from | One or more origin papers | A DOI, OpenAlex id or title; up to 10 Library papers; or the papers of a Deep Research answer |
| Graph | Similarity layout, clusters | Years across, circle size by citations, seeds and set papers colored |
| Lists | Prior works, Derivative works [3] | References, Citing, Snowball and Reviews tabs, filterable by year, citations and study type |
| Export | Not confirmed on its pages | RIS, CSV (with round), Save to Library |
| Alerts | Not described on its pages | None |
| Free use | Monthly graph cap, reported as 5 graphs a month [2] | Unlimited maps at 0 credits with a free account (fair-use rate limit) |
| Paid plans | Academic and Business plans [2] | Not needed for Citation Map; Plus $12/month and Pro $29/month cover AI work |
How Citation Map works, step by step
- Start a map. Open Citation Map, paste a DOI or paper title and click Map it. Or click From my Library to start from up to 10 saved papers, or open it from a Deep Research result.
- Read the map. The Map tab shows your seed in orange, its references and citing papers placed by year, and circle size by citation count. For one seed, the References and Citing tabs list up to 40 papers each, most-cited first.
- Build your set. Click the plus next to any paper to add it. The "Your set" panel lists every paper with its round.
- Run a round. Choose Backward, Forward or Both. Candidates linked to several of your set come first. Leave Hide papers already seen on so each round shows only new papers.
- Check for reviews. Click Find reviews citing the set to list reviews and meta-analyses that cite your papers.
- Export. Download RIS or CSV, or Save to Library.
Example: a master's thesis on sleep and exam scores
Example (invented, for illustration). A student has one strong cohort study on sleep duration and exam performance. In Connected Papers she sees two clusters, sleep duration and chronotype, and notes 6 candidates. She moves to Citation Map, adds the original paper and the 6 candidates to her set, and runs a Both round: of 60 ranked candidates, 14 are linked to 3 or more set papers, and she keeps 9 after reading abstracts. Find reviews citing the set returns 4 reviews; one lists 2 studies she had not seen. She exports the 18-paper set as CSV and writes: "Backward and forward citation searching from 7 seed papers (one round, November 2026) identified 11 additional studies."
Which should you pick?
| If your job is to... | Pick |
|---|---|
| See the shape of a new field in a minute | Connected Papers |
| Find papers similar to yours that never cite each other | Connected Papers |
| Run documented backward and forward rounds for a thesis or review | Citation Map |
| Find reviews and meta-analyses that cite your key papers | Citation Map |
| Move found papers straight into screening, reading and writing | AI for Research |
| Get alerts about new papers over months | Litmaps or ResearchRabbit; see Litmaps alternative and ResearchRabbit alternative |
How to use them together
The two tools cover different parts of the same search, so many people use both.
- Use Connected Papers once at the start to find the clusters and the classics.
- Pick 3 to 10 central papers and start a Citation Map from them.
- Run one or two rounds, screen candidates against your criteria, and stop when a round adds nothing new.
- Export the set as RIS and upload it into a Systematic Review, or save it to your Library to read and write with it.
Record both in your methods: similarity mapping is exploration; the rounds are your citation search.
What Citation Map does not do
- No similarity layout. It shows direct citation links. Papers that share a topic but no links will not appear, which is exactly where Connected Papers is stronger.
- No alerts. It will not tell you about new papers next month.
- Set limits. A map starts from up to 10 seeds, and each snowballing round reads the first 25 papers of your set and returns up to 60 candidates.
- Coverage follows OpenAlex. Books, some humanities work and papers from the last few weeks are thinner. Run a keyword search in Find Papers or a library database first.
- Maps are personal. A map is saved to your History; you share an export of the set, not a live map link.
- Sign-in needed. Citation Map is free but needs an account.
Who it is for
Master's and PhD students who must describe their search in a methods chapter, and review teams who report citation searching as an "other method". See AI for PhD students and AI for a master's thesis.
Learn more
- Connected Papers vs ResearchRabbit vs Litmaps: the three-way comparison.
- Citation snowballing: the method, and when to stop.
- How to find who cited a paper: the quick version for one paper.
- Research Rabbit alternatives: seven discovery tools in one list.
Sources
- Connected Papers. About. https://www.connectedpapers.com/about (checked November 2026; page content did not load for our checker, method as quoted in library guides)
- Connected Papers. Pricing. https://www.connectedpapers.com/pricing (checked November 2026; plan details did not load for our checker)
- Loyola Marymount University Library. AI research tools: Connected Papers. https://libguides.lmu.edu/AIresearchtools/CP (checked November 2026)
Last updated 2026-11-12