Rayyan
Web-based systematic review screening platform — streamlines title/abstract screening with AI-assisted relevance suggestions and blind collaboration between reviewers, widely used for PRISMA-compliant reviews.
What it does
Rayyan is a web application designed specifically for the title/abstract screening stage of a systematic literature review. After you export search results from Scopus, PubMed, Web of Science, or other databases, you import them into Rayyan, which presents each record for inclusion/exclusion decisions and tracks reviewer decisions with reasons.
Key capabilities:
- AI-assisted screening suggestions: Rayyan’s ML model learns from your early decisions and suggests relevance scores for unscreened records, allowing you to prioritize or batch-exclude lower-priority records
- Blind collaboration: multiple reviewers can screen the same records independently without seeing each other’s decisions — the system then surfaces conflicts for reconciliation
- PRISMA flow tracking: automatically tracks how many records were identified, screened, excluded, and included at each stage
- Conflict resolution: highlights records where reviewers disagreed for adjudication
- Multiple import formats: accepts RIS, CSV, and direct imports from major databases
Best for
Systematic and scoping reviews that involve large volumes of records (hundreds to thousands) requiring title/abstract screening with documented inclusion/exclusion decisions. Especially useful for multi-reviewer teams where blind screening and conflict resolution tracking are required by the review protocol.
Pricing
Freemium. The free plan covers a limited number of active reviews and records. The paid plan (Rayyan Premium, ~$10/month per reviewer) removes limits and adds advanced collaboration and export features. Many institutions have site licenses — check with your library.
Strengths
- Purpose-built for the screening step: the interface is optimized for fast include/exclude decisions with keyboard shortcuts, reducing screening time per record
- AI suggestions accelerate screening: by learning from your early decisions, the model can flag likely-irrelevant records for batch review, reducing time on clearly off-topic records
- Blind collaboration is handled natively — no need to coordinate who sees what across spreadsheets
- PRISMA-compatible flow diagram is generated automatically from your decisions
- Works with records from any database that exports RIS or CSV
- Free tier is sufficient for a single review with typical record volumes
Limitations
- AI suggestions are a starting point, not a substitute for reviewer judgment — over-relying on relevance scores can introduce bias; always document your screening approach
- Free plan limits the number of concurrent active reviews, which matters for research groups running multiple reviews simultaneously
- Rayyan focuses on title/abstract screening only — full-text screening and data extraction happen outside the tool (typically in Elicit, spreadsheets, or dedicated software like Covidence)
- The collaboration model assumes all reviewers have Rayyan accounts; large teams coordinating across institutions may encounter friction
How it compares
| vs. | Key difference |
|---|---|
| Elicit | Elicit extracts structured data from full texts; Rayyan handles title/abstract screening. They address sequential stages — use Rayyan first, then Elicit on included papers |
| Covidence | Covidence is the premium alternative to Rayyan, with better full-text screening and data extraction; it’s more expensive (institutional licensing). Rayyan is the more accessible free option for title/abstract screening |
| Manual spreadsheet screening | Spreadsheets don’t support blind review, conflict tracking, or AI suggestions; Rayyan is strictly better for any systematic review with more than one reviewer |
Related content
- Tutorial: Running a Systematic Literature Review with AI
- Tool: Elicit — for full-text data extraction after Rayyan screening
- Tool: Scopus — for generating the search results you import into Rayyan
- Tool: Zotero — for managing your final included reference set