Defensible literature review blog banner
Tool ComparisonBy Cicadus TeamApril 2026

Cicadus vs Scite vs Connected Papers: An Honest Comparison from a Solo Builder

I built Cicadus. I've used both Scite and Connected Papers extensively. Here's my honest take on which tool actually fits your research workflow — including where Cicadus falls short.

When I started building Cicadus, the first thing I did was spend two weeks using every research tool I could find. Not to copy them — to understand what they were good at and where the gaps were.

Scite and Connected Papers kept coming up. Researchers love them. And for good reason — they solve real problems. But they solve different problems than Cicadus does.

This comparison is honest. I'll tell you where each tool wins, where it loses, and exactly when you should use one over the other. I'll also be upfront about where Cicadus isn't the right choice.

The short version: Scite tells you if a paper is supported or contradicted by later research. Connected Papers shows you what a research field looks like visually. Cicadus tells you why papers cite each other — and finds papers you didn't know to look for. They're complementary, not competing.

What Each Tool Actually Does

Before comparing, it's worth being clear about what problem each tool is solving — because they're genuinely different.

ToolCore question it answersPrimary mechanism
SciteHas this paper been supported or contradicted?Smart Citation classification at scale
Connected PapersWhat does the landscape around this paper look like?Visual graph via co-citation and bibliographic coupling
CicadusWhy do papers cite each other — and where do they disagree?Citation intent classification + semantic search

Scite — Smart Citations at Scale

Scite.aiFree (limited) · $20/month individual

Shows you whether papers support or contradict each other

Scite's core insight is brilliant: not all citations mean the same thing. When paper A cites paper B, is A saying "this paper proves our point" or "this paper got it wrong"? Scite classifies citations into three categories — supporting, contrasting, and mentioning — and has indexed over 1.6 billion citation statements to do this at scale. If you want to know whether a specific claim has been validated or challenged by subsequent research, Scite is exceptional. It's particularly powerful for medical and biological research where the credibility of a study genuinely matters. The browser extension lets you see citation classifications anywhere you're reading online.
What it does well
  • 1.6B+ indexed citations
  • Browser extension for any site
  • Reference check for manuscripts
  • Strong in medicine and biology
  • AI assistant for synthesis
Limitations
  • $20/month is expensive for students
  • Only 3 citation categories
  • Weaker coverage in CS and humanities
  • No semantic search for related papers
  • Free tier severely limited

When Scite is the right choice: You have a specific paper and want to know how the academic community has received it. You're doing systematic reviews and need to validate whether your key references have been contradicted. You're in medicine, biology, or life sciences where citation credibility is critical.

Connected Papers — The Field Map

Connected PapersFree (5 graphs/month) · $6/month academic

Shows you the visual landscape of a research field

Connected Papers solves a different problem entirely. You have one paper you know is relevant — but you don't know what else is out there. Connected Papers takes that seed paper and generates a visual graph of related work using co-citation and bibliographic coupling signals. Papers that are cited together or cite the same sources cluster together. The result is a map of a research field. You can see which papers are foundational (large nodes, heavily cited), which are recent, and how ideas cluster. It's one of the best tools for the "I don't know what I don't know" phase of research.
What it does well
  • Intuitive visual graph
  • Discovers papers keyword search misses
  • Prior and derivative works view
  • Free tier generous for most needs
  • Multi-origin graphs available
Limitations
  • Only covers Semantic Scholar index
  • No citation intent classification
  • 5 graphs/month on free tier
  • No semantic search by concept
  • Shows connections, not their meaning

When Connected Papers is the right choice: You're entering a new field and need to understand the landscape quickly. You want to find the foundational papers everyone builds on. You work better visually and want to see how ideas connect before diving into reading.

Cicadus — Citation Intent and Conflict Detection

CicadusFree · Pro US$6.99/mo (₹299)

Tells you why papers cite each other — and where they disagree

Cicadus was built to answer a question the other tools don't address well: not just what the literature says, but where it fights. When you're doing a literature review, knowing that paper A cites paper B is less useful than knowing paper A cites paper B to challenge its methodology. Cicadus classifies citation intent into four categories — background, methodology, critique, and application — and uses semantic search to find related papers based on meaning rather than keywords. Papers are retrieved from major scholarly indexes and paper banks (not a single in-house corpus). Researchers have pointed out that the conflict and disagreement detection is the standout feature: you can filter a citation network to see which papers actively contradict each other, surfacing debates that a surface-level reading would miss entirely.
What it does well
  • 4-category citation intent classification
  • Conflict and disagreement detection
  • Semantic search — finds by meaning
  • Visual citation networks
  • Free tier for core workflows; Pro unlocks higher limits
Being honest about limitations
  • Not a billion-row proprietary citation index like Scite
  • Less established than both competitors
  • Coverage depends on upstream sources and field
  • Fewer integrations — still early
  • No browser extension yet

When Cicadus is the right choice: You're doing a literature review and need to understand not just what papers exist but how they relate conceptually. You want to find contested claims and active debates. You're searching by concept and keep missing relevant papers. Or you're a student and $20/month for Scite isn't feasible.

Head-to-Head Comparison

FeatureCicadusSciteConnected Papers
Citation intent classification✓ 4 categoriesPartial (3 types)None
Conflict / disagreement detectionPartial
Semantic search (by concept)
Visual citation networkLimited✓ Best in class
Database / sourcesPapers retrieved from major scholarly indexes and paper banks (multi-source)1.6B+ citations (owned index)Semantic Scholar
Free tier✓ US$0 — core tools; usage caps applyVery limited5 graphs/month
Paid plan costPro US$6.99/mo · ₹299/mo (India)~$20/month~$6/month academic
Browser extensionNot yet
Best forUnderstanding citation relationships and debatesValidating specific claims and credibilityExploring a field visually from one paper

Which Tool for Which Situation

Starting a literature review from scratch
Start with Connected Papers. Enter your most relevant seed paper. Get the field map. Identify the key papers. Then use Cicadus to understand how those papers relate conceptually and where debates exist.
Checking if a key reference is still credible
Use Scite. If you're citing a specific study in your own paper, Scite tells you whether it's been contradicted by newer research. This is what Scite is uniquely built for. Cicadus and Connected Papers don't replace this.
Finding contested claims and active debates
Use Cicadus. Filter citation networks by critique and conflict intent. See which papers actively challenge which others. Several researchers have described this as the feature they didn't know they needed.
You're a student on a budget
Use Cicadus and Connected Papers. Cicadus's free tier covers core research workflows; Pro is optional when you need higher limits. Connected Papers' free tier covers most graph needs. Scite's free tier is severely limited — full access is often ~$20/month without institutional access.
Systematic review for medical or clinical research
Use Scite as primary. Its 1.6B+ citation database and focus on supporting vs contrasting evidence is purpose-built for systematic reviews in medicine and biology. Use Cicadus as a complement to find semantically related papers you might miss.

The Workflow That Uses All Three

These tools aren't competing — they cover different parts of the research process. Here's the workflow several researchers have described to me:

Step 1 — Field exploration (Connected Papers)Enter your seed paper. Identify the 5–10 most important papers in the field. Understand the landscape before reading anything deeply.
Step 2 — Semantic discovery (Cicadus)Take those key papers into Cicadus. Use semantic search to find conceptually related papers. Filter the citation network by intent to see which papers are critiquing each other and where active debates exist.
Step 3 — Credibility check (Scite)Before citing a key study, run it through Scite's Reference Check. Confirm it hasn't been significantly contradicted by newer research. Takes 5 minutes and can save you from citing a disputed study.
The honest bottom line: If you can only use one tool, use whichever fits your specific task. Exploring a field visually — Connected Papers. Validating specific citations — Scite. Doing a literature review and wanting to understand the debates — Cicadus. Student on a budget — Cicadus and Connected Papers get you most of the way there on their free tiers.

This comparison was written by Rakesh, who built Cicadus as a solo project. Feedback from researchers actively shapes what gets built next — including the conflict detection feature, which came directly from a Reddit comment.

Map the citation structure of your field at cicadus.com

Start exploring Cicadus →

Further Reading

Scite.ai — scite.ai · Smart Citations platform

Connected Papers — connectedpapers.com · Visual literature mapping

Boote & Beile (2005) Scholars Before ResearchersEducational Researcher, AERA

Webster & Watson (2002) Analyzing the Past to Prepare for the Future: Writing a Literature Review MIS Quarterly