Elicit

Free tier

AI-powered scientific research assistant that helps you find, analyze, and synthesize academic literature

Free tier available·All audiences·Powered by Anthropic (Claude)·API available

Key strengths

Searches 138M+ academic papers and 545K+ clinical trials with semantic searchAutomates systematic literature reviews with up to 80% time savingsSentence-level citations for full transparency and traceabilityPRISMA 2020-compliant systematic review workflow with reproducible auditingHigh accuracy: 95% search recall, 97% abstract screening, 99% full-text screening

Free tier + paid plans · from $11 USD/mo

Facts verified against elicit.com on Oct 8, 2026

Recent changes

Oct 8, 2026Plans updated: Free Basic plan; Plus at $11/user/month (billed $132/yr); Pro at $39/user/month (billed $468/yr); Scale at $89/user/month (billed $1,068/yr); Enterprise at custom pricing. Monthly pricing also available (Pro $49/mo, Scale $169/mo).

What's changed across AI tools →

Get alerts when Elicit changes

Free. We'll email you when pricing, the free tier, API access or open-source status changes.

No ratings yet

Technical Documentation & Integration

Elicit provides developer-facing capabilities for embedding AI-powered research into custom workflows:

API access — available for programmatic querying of Elicit's research capabilities
MCP server — supports integration with Model Context Protocol-compatible agents, enabling autonomous research pipelines
SSO & SAML — available on Enterprise plans for organizational identity management
Custom data sources — Enterprise plans support ingestion of proprietary or domain-specific corpora
Export formats: RIS, CSV, BIB, PDF, DOCX — enabling downstream processing in reference managers, data pipelines, or document workflows
Zotero integration — direct sync with the Zotero reference manager

Developers building research automation or RAG pipelines should note Elicit's sentence-level citation architecture, which makes AI outputs verifiable at the claim level — critical for compliance-sensitive or high-stakes research environments. The platform can process up to 20,000 data points per run, making it viable for large-scale extraction and synthesis tasks.

Alternatives