Connected Papers Ltd.’s visual academic research graph. Paste a paper by title, DOI, URL, or ArXiv ID and get a similarity graph showing related papers arranged by co-citation and bibliographic coupling, not keyword matching. Center nodes are foundational. Periphery nodes are newer or more specialized.
Built on Semantic Scholar’s index of 200M-plus papers.
Pricing: Free (5 graphs/month), Academic $3/mo billed annually for unlimited.
System Verdict
Pick Connected Papers when you know one landmark paper but not the surrounding landscape. The graph surfaces foundational works and recent developments that keyword search misses entirely. Best use case is the first 30 minutes of scoping a new literature review.
Skip it when you need synthesis, Q&A, or citation sentiment. Connected Papers finds papers. It does not read them. Elicit extracts structured data across hundreds of studies. Consensus synthesizes answers from abstracts. Scite classifies citations as supporting or contrasting.
Who pays which tier: Free for occasional mapping (5 graphs/month covers casual use), Academic $3/mo billed annually for any PhD student or researcher running multiple reviews. Cheapest upgrade in the research-tool category.
Key Facts
| Seed input | Title, DOI, URL, or ArXiv ID |
| Underlying index | Semantic Scholar (200M+ papers) |
| Graph algorithm | Co-citation and bibliographic coupling, not keyword match |
| Prior Works view | Foundational papers the seed builds on |
| Derivative Works view | Work that built on the seed paper |
| Free tier | 5 graphs/month, no account required |
| Academic tier | $3/mo billed annually, unlimited graphs, export |
| Export | Graph image and paper metadata |
| Coverage ceiling | Inherited from Semantic Scholar |
What it actually is
One web tool that does one thing: build a similarity graph from a seed paper. Nodes are papers sized by citation count. Edges imply shared references and co-citation patterns. Hover shows title, authors, year, and abstract snippet. Clicking any node opens it on Semantic Scholar or jumps to a new graph centered on that paper.
The Prior Works view ranks foundational papers the seed builds on. The Derivative Works view tracks what came after. Both reveal a field’s structure at a glance, something keyword search cannot do.
The moat is thin. The similarity algorithm is proprietary but the underlying data is Semantic Scholar’s free index. ResearchRabbit offers similar graph exploration with added collaboration features. Connected Papers wins on speed and single-graph focus. ResearchRabbit wins on team workflows.
When to pick Connected Papers
- You are starting a literature review in an unfamiliar field. One DOI reveals the conceptual neighborhood faster than any keyword search.
- You are writing a grant’s related-work section. Map the landscape, identify foundational citations, avoid missing obvious priors.
- You learn visually. The graph shows clustering and influence in a way a citation list cannot.
- You need to verify a paper sits in the right neighborhood. Papers that cluster unexpectedly often signal methodological overlap worth checking.
- You run multiple reviews per month. Free tier caps at 5 graphs. Academic at $3/mo unlocks unlimited.
When to pick something else
- Cross-paper synthesis or research Q&A: Consensus or Elicit. Connected Papers does not read or reason across papers.
- Structured extraction for meta-analysis: Elicit. Pulls sample sizes, outcomes, and effect sizes from hundreds of papers.
- Citation sentiment (Supporting vs Contrasting): Scite. Classifies the 1.2B+ citations Connected Papers only counts.
- Free comprehensive search: Semantic Scholar. Same underlying index, full search interface, TLDR summaries, free API.
- Team collaboration on literature maps: ResearchRabbit offers richer team features at the trade-off of speed and focus.
Pricing
| Plan | Price | Graphs/Month | Features |
|---|---|---|---|
| Free | $0 | 5 graphs | Basic graph, Prior/Derivative Works, no account required |
| Academic | $3/mo (billed annually) | Unlimited | Batch operations, graph export, metadata export |
| Business | Custom | Unlimited | Team features, institutional licensing |
Prices verified 2026-05-13 via Connected Papers pricing. Free tier requires no signup. Academic billed annually.
Against the alternatives
| Connected Papers | Semantic Scholar | ResearchRabbit | |
|---|---|---|---|
| Primary output | Visual similarity graph | Paper list + TLDR | Graph + collaboration |
| Underlying corpus | Semantic Scholar (200M+) | Own index (200M+) | Semantic Scholar + others |
| Free tier | 5 graphs/month | Fully free, forever | Free with account |
| Price floor (paid) | $3/mo Academic | $0 | Free |
| Team features | Minimal | None | Built-in |
| Graph speed | Fast | N/A | Slower |
| Best viewed as | Solo field-mapping tool | Free discovery layer | Team literature workflow |
Failure modes
- Finds, does not read. Connected Papers surfaces related work. It will not extract data, answer research questions, or synthesize findings.
- 5 graphs/month free tier is thin. A single literature review often consumes it. Active researchers will hit the cap in week one.
- No citation classification. Cannot distinguish supporting from contrasting citations. Scite fills that gap.
- Coverage ceiling inherited from Semantic Scholar. Humanities and some social sciences are less comprehensively indexed.
- Graph can overwhelm for popular seed papers. Well-cited works in large fields produce dense graphs with limited filtering.
- No synthesis or chat layer. Researchers who want both a graph and Q&A will need a second tool.
- Single seed paper required. Multi-origin graphs exist but still start from user-selected roots, not free-form research questions.
Methodology
This page was produced by the aipedia.wiki editorial pipeline, an automated system that ingests vendor documentation, verifies pricing and feature details against primary sources, and generates the analysis above. No individual human wrote this review. Scoring follows the four-dimension rubric at /about/scoring/ (Utility × Value × Moat × Longevity, unweighted average). Last verified 2026-05-13 against Connected Papers pricing and Semantic Scholar about.
FAQ
How is Connected Papers different from Google Scholar citation tracking? Google Scholar shows direct citations. Connected Papers shows conceptual similarity using co-citation and bibliographic coupling. Papers can share themes and references without citing each other directly, which the graph reveals.
Does Connected Papers work for all fields? Coverage strongest in STEM, computer science, and social sciences. Humanities and arts are sparser because the underlying Semantic Scholar index is thinner there.
Is the Academic plan worth $3/month? For active researchers, yes. The 5 graph/month free cap disappears fast during a real literature review. Academic is the cheapest productivity upgrade in the research-tool category.
Can I use Connected Papers offline? No. All graph generation runs server-side against Semantic Scholar’s index.
Connected Papers vs Elicit: which one? Different jobs. Connected Papers maps the conceptual neighborhood from one seed paper. Elicit runs a structured systematic review across hundreds of papers. Most researchers use Connected Papers first to scope, then Elicit to extract.
Does Connected Papers support team collaboration? Minimal. Single-user workflow by design. Teams needing shared literature libraries should look at ResearchRabbit or institutional tools like Covidence.
Sources
- Connected Papers pricing: current tier prices and limits
- Semantic Scholar about: underlying index size and coverage
- Connected Papers Premium announcement: feature history
Related
- Category: AI Research
- Comparisons: Connected Papers vs Semantic Scholar | Connected Papers vs Elicit | Connected Papers vs Consensus | Connected Papers vs Scite
