Elicit capabilities

Semantic search across over 138 million academic papers aggregated from Semantic Scholar, PubMed, and OpenAlex. Finds relevant research even when exact keywords are absent, returning up to 1,000 papers per query with relevance scoring.

Fully automated research overviews generated from a natural-language research question. Elicit searches papers, screens them, extracts data, and synthesizes findings into a structured report with sentence-level citations. Reports cover up to 200 papers (changelog '200 Paper Reports', 7 February 2026).

Users define custom columns—each with a name and natural-language extraction instructions—to pull specific data points from paper full text or tables. Supports open-ended, yes/no/maybe, and multiple-choice column types.

A step-by-step workflow covering search, title-and-abstract screening, full-text screening, and data extraction for systematic literature reviews. Supports dual review, PRISMA 2020 guidelines with exclusion reasons, per-criterion scores, and supporting quotes for auditability.

A single-page workspace that unifies search, summaries, and systematic review. Users run multiple queries on one page, combine papers across queries, chat with selected papers, and create flexible paragraph summaries.

A REST API that exposes Elicit's core search and synthesis capabilities for integration into external tools and scripts. Supports semantic paper search, clinical trials search, automated report generation, and systematic review pipelines. Initially launched in preview (March 2026); as of June 2026 the docs.elicit.com reference does not carry a preview qualifier.

An autonomous agentic workflow that takes a natural-language research question, breaks it down into a systematic research program, and executes that program end to end to produce a structured deliverable where every claim is grounded in evidence. It powers higher-level workflows than a single literature search - Competitive landscapes, Research landscapes, Clinical trial analyses, and open Topic exploration - and lets users iterate on sources and outputs until the result matches their need. As of April 2026 the agent can assemble tables, figures, and slides directly into its output, reflecting the 2026 shift toward agentic research assistants rather than search-and-summarize tools.

Saved natural-language research questions that automatically surface new papers as they are published. Each alert assigns relevance scores to new papers and optionally sends email notifications.

Every AI-generated claim—in summaries, data extractions, Reports, and Research Agent outputs—is backed by an exact sentence or figure quote from the underlying paper. This is applied uniformly across all surfaces rather than being an opt-in toggle.

Elicit offers four plan tiers with increasing quotas for reports, systematic reviews, custom columns, alerts, and API access. Since the July 2024 pricing overhaul, paper search, summaries, and chat with papers are free and unlimited on all tiers and credits were phased out; gating now applies to Reports/Systematic Reviews per year, extraction columns, alerts, screening scale, and API. The public pricing page (verified 2026-06-07) lists exactly four purchasable tiers: Basic (free), Pro, Scale, and Enterprise. Earlier 'Plus' and 'Team' tiers are no longer offered on the pricing page (Elicit's help-center articles still reference a legacy 'Team' tier, but it is not purchasable on the live pricing page).