Scite.ai

Scite.ai is an AI research platform built around Smart Citations, a proprietary classification system that analyzes how scientific papers are cited across academic literature. Rather than providing citation counts, Smart Citations classify each citation statement as supporting, contrasting, or mentioning the cited work, giving researchers immediate signal on whether findings are validated, contested, or simply referenced in subsequent literature. The platform indexes 1.2 billion+ citation statements and 280M+ full-text articles through licensing agreements with major publishers. Features include the Scite Assistant (AI Q&A grounded in scientific literature), full-text search, custom dashboards for tracking papers and authors, citation alerts, journal and institutional analytics, and an official MCP server for AI agent integration. Founded in Brooklyn in 2018 and funded in part by the National Science Foundation and the National Institute on Drug Abuse (NIH), Scite.ai serves academic researchers, pharmaceutical scientists, science journalists, and research-intensive organizations globally. Scite.ai offers meaningful integration depth across several channels. An official MCP server (live at scite.ai/mcp, listed on PulseMCP) enables AI agents including Claude, ChatGPT, and other MCP-compatible clients to access Smart Citation analysis programmatically across 280M+ full-text articles. A Zotero plugin enables direct citation analysis within the leading open-source reference manager. A browser extension surfaces Smart Citation context while browsing publisher sites. Organization-tier API access supports institutional and programmatic integration. Publisher data agreements provide full-text PDFs and XML directly from major publishers for citation classification. Named integration gaps: no Mendeley native connector, no Overleaf or LaTeX integration for manuscript workflows, and API access is gated to the Organization tier and not available on the Personal plan. Personal pricing is publicly listed. Personal: $20/month billed monthly or $12/month billed annually (40% saving, $144/year). A 7-day free trial of the Personal plan is available with no permanent free tier. Organization tier: custom pricing for companies, universities, and research institutions, including everything in Personal plus flexible access via domain, IP, or email-based licensing, Organization API, tiered volume pricing, user analytics, additional datasets (Patents, Clinical Trials), advanced AI models, shared dashboards, and enhanced data exports. Organization pricing is not publicly listed and requires contacting the sales team directly. Scite.ai is not suited for researchers who need comprehensive literature synthesis beyond citation context. The platform classifies citations but does not autonomously conduct multi-round adaptive literature searches, generate full annotated research reports, or extract structured data across multiple studies. For depth-first citation trail search with full annotated reports, Undermind ($16/month annual) covers that workflow. For AI-generated paper summaries and structured data extraction, Elicit ($10/month) provides a dedicated systematic review pipeline. For open-domain research beyond peer-reviewed literature, Gemini Deep Research (5 reports/month free, $19.99/month Pro) or ChatGPT Deep Research ($20/month) cover broader source territory. For basic citation lookup without AI analysis, Semantic Scholar (free) provides comparable coverage at no cost. Trustpilot reviewers note weaker performance on complex synthesis tasks compared to Perplexity and deep research tools; Scite.ai is strongest at citation context classification, not general-purpose research generation. Scite.ai launched its official MCP server at scite.ai/mcp in 2025, enabling AI agents to access Smart Citation analysis programmatically and positioning Scite as a citation-aware data layer for agentic research systems. The 7-day free trial remains available with no permanent free tier as of Q2 2026. Trustpilot shows 222 reviews with a 4-star TrustScore; G2 shows 27 reviews at 4.7/5. Negative Trustpilot feedback documents AI hallucinations in the Scite Assistant including fabricated citations with DOI links, broken BibTeX export functionality, UX quality issues, and customer support response time concerns. Coverage is strongest for indexed publisher literature; preprint coverage and niche field coverage have gaps noted by users. NSF and NIH funding provides institutional stability alongside commercial operations.