Elicit

  • Elicit — AI Research Assistant for Systematic Reviews & Structured Extraction (Ought Spin-out, Active, Free Tier + $49 Pro Floor)

    Not shut down.​ Elicit launched 2023, spun out of Oakland nonprofit Ought (CEO Andreas Stuhlmüller), raised $22M Series A Feb 2025 ($33M total), 2M+ researchers, NASA/NIH/CDC/Harvard/Yale/J&J on logo wall. Active 2026-08, independent, no acquisition/rebrand/shutdown. Corpus = 138M+ papers + 545K ClinicalTrials.gov. Category ai-productivity-data; pipeline_stage none. Killer differentiator = sentence-level cited structured extraction into custom columns (sample size / effect size / adverse events) + guided 4-step PRISMA-style Systematic Review workflow (search→screen 5,000→extract→report) + Research Agent that screens and tables without per-paper human reading; not Consensus (yes/no evidence meter, no extraction tables), not Scite (citation support/contradict context only), not Semantic Scholar (discovery graph, no extraction), not ChatGPT Deep Research (web-general, no PRISMA screen, hallucinates DOIs). 2026 trap: free = 2 reports/mo + 2 columns only (real SLR needs Pro $49), recall ~39.5% sensitivity in 2025 PMC study vs ~94.5% human (must supplement PubMed/Scopus/Web of Science), extraction needs human verification on critical arms, biomedical/social-science weighted, humanities/law thin, no web/news source, Plus $12 tier quietly sits between Free and Pro on some pages.

    1. One-line positioning

    Elicit = systematic-review engine, not chat-with-PDF: plain-English question (“interventions reducing HF readmission?”) → semantic search 138M papers + ClinicalTrials.gov → screen up to 5,000 by inclusion rules → auto-populate custom-column table (method, N, outcome, effect size, limitations) with every cell linked to source sentence​ → Research Agent runs 10+ page cited Report → export CSV/BIB/RIS → Zotero sync → Research Alerts ping new matches. Fits ai-productivity-data; pipeline_stage none (query→retrieve→screen→extract→table→report→export). For PhD/meta-analysis/pharma HTA/policy briefs, it compresses screening+extraction from weeks to hours; for “quick is X true?” Consensus/Perplexity cheaper, for cite-trust Scite, for discovery Semantic Scholar+ResearchRabbit.

    2. Core features

    Layer Capability Notes
    Corpus 138M+ papers (Semantic Scholar/OpenAlex/PMC), 545K ClinicalTrials.gov, PubMed bridge Biomed/social/econ strongest; engineering/humanities/law thinner
    Semantic search Meaning-based, not keyword; returns relevant studies with match reason
    Systematic Review workflow 4-step guided: search → screen (up to 5,000 Pro / 40,000 Ent) → extract → report; PRISMA-trackable Pro unlocks full workflow
    Structured extraction Custom columns (20 Pro / 30 Scale / 40 Ent), each value sentence-cited, table to CSV/BIB/RIS/XLSX Extraction accuracy >90% reported, 99.4% in one German Ed policy SLR; still verify critical cells
    Research Agent Autonomous screen+extract+report across a paper set; 2/mo free, 144/yr Pro, 240/yr Scale
    Elicit Reports 10+ page cited brief across up to ~80 papers, methods+findings+limits Free 2/mo, Plus 4, Pro 144/yr
    Chat with Papers Ask a paper/set questions, ground to passages; 8 full-text at once (Pro) OA-only unless you upload PDF or browser-ext leverages your subscription
    Figure extraction Pull values from figures/tables in PDFs Scale+ only
    Research Alerts AI-filtered new-paper alerts (10 Pro / unlimited Ent), not noisy keyword ping
    Integrations Zotero import/export, Mendeley/EndNote export, REST API (Pro+), official MCP server​ (Claude Desktop/Code) SOC 2 Type II
    Notebook Collect papers, multi-query synthesis

    3. Pricing (2026-08, USD)

    Tier Monthly Annual Caps
    Basic (Free)​ $0 $0 Unlimited search + summaries + paper chat (OA), 2 reports/mo, 2 columns, limited Research Agent, Zotero import, no alerts, no API
    Plus​ (some pages) $12​ $120​ 4 reports/mo, expanded chat, no custom columns, no systematic workflow, no alerts
    Pro​ $49​ $499​ ($588 on some pages, 35% off) Systematic Review (5,000 screen), 20 columns, 144 reports/yr, 10 alerts, API, custom extraction, 8-paper chat
    Scale​ $169​ $2,028​ 5× Pro usage, figure extraction, 30 columns, 240 reports/yr, live collaboration, admin/seat
    Team​ $79/seat​ $780/seat​ Pro features, pooled reports, shared notebooks, 2-seat min
    Enterprise​ Custom Custom 40,000 screen, 40 columns, SSO/SAML, no-train-on-your-data, dedicated CS

    pricing_type = freemium-then-per-seat-usage-metered-reports; starting_price = $0; free_tier = "yes; unlimited search/summaries/paper-chat, 2 automated reports/mo, 2 extraction columns, limited Research Agent, NO systematic-review workflow/alerts/API/custom-columns"; pricing_note = "2026-tiers-basic-free-plus-12-pro-49-499-scale-169-enterprise; pro-588-annual-on-official-page-499-on-some-resellers-verify-checkout; reports-are-meter-free-2-plus-4-pro-144yr-scale-240yr; screen-5000-pro-40000-ent; plus-tier-between-free-pro-confuses-buyers; api-pro-plus-mcp-server-live; extraction-cells-need-human-verify-critical-arms; recall-39.5pct-sensitivity-2025-pmc-study-supplement-pubmed-scopus"

    4. Access Type

    access_type: web-dashboard-plus-api-plus-mcp

    access_display: 🌐 elicit.com · 🔌 Zotero/Mendeley/EndNote · REST API (Pro+) · MCP server (Claude Desktop/Code, 2026) · browser ext for subscriber full-text · no native iOS/Android app

    5. Reviews (2026)

    Source Read
    AI Agent Index 4.7/5 — “strongest institutional validation of any AI research assistant; NASA/CDC/Harvard logos; MCP+REST+SOC2”
    Independent (PhD evals) Reports beat ChatGPT/Perplexity/Gemini Deep Research on time-savings + citation transparency
    2025 PMC/medRxiv study Sensitivity only ~39.5%​ vs ~94.5% human review; precision ~40% — “powerful assistant, not sole search method”
    G2/Capterra Thin (0–1 reviews, unclaimed) — procurement must rely on case studies not peer reviews
    ToolCompare “Pro $49 is 4–5× Consensus/SciSpace; steep wall for casual”

    Praise: sentence-level citations per cell (auditable), custom-column extraction replaces weeks of manual tabulating, PRISMA workflow built-in, ClinicalTrials.gov alongside papers rare, Research Agent screens 5K unattended, MCP+API enterprise-ready, free tier usable for exploration.

    Gripes: free 2 reports/mo vanishes in thesis crunch, recall incomplete (must cross-search PubMed/Scopus/WoS), extraction misreads table/confuses study arms on hard cases, biomedical-weighted (theory/humanities/law thin), OA-only unless upload/ext, no web/news/business source, Plus tier naming inconsistent across pages, G2 empty so social proof is logo-wall not peer-review.

    6. Best for / Not for

    Best for Not for
    PhD/dissertation/meta-analysis authors running PRISMA SLRs Casual “is X true” lookups (Consensus $20/Perplexity $20)
    Pharma/medtech/HTA teams building trial-evidence tables by arm Humanities/theory-heavy or legal-history research (corpus thin)
    Policy/government analysts writing evidence briefs on deadline Teams needing web/news/patent/market sources (pair Perplexity)
    Systematic-review labs co-authoring extraction tables (Scale) People who won’t verify extractions against PDFs (compliance risk)
    Researchers wanting MCP into Claude Code workflow General writing/coding (ChatGPT/Claude better)

    7. Competitors

    Tool Lane Starts
    Consensus Evidence-meter yes/no synthesis, no extraction tables Free / $20-mo
    Scite Smart citations (support/contrast/mention), no SLR workflow $20/mo
    Semantic Scholar Free 200M discovery + citation graph, no extraction Free
    ResearchRabbit / Connected Papers Citation-network mapping, free Free
    SciSpace (Typeset) Chat-with-PDF + explain, lighter extraction $20/mo
    NotebookLM Synthesize your uploaded source pack Free / $4.99
    ChatGPT/Claude Deep Research Web-general long reports, hallucinates DOIs $20/mo
    Covidence / Rayyan / DistillerSR Traditional PRISMA review mgmt, less AI extract $20–85/mo

     

Do Not Sell or Share My Personal Information Cookie Settings