Overview
Elicit automates literature review: search across a research corpus, extract structured data from papers into tables, and summarize findings across sources. Search and summaries are effectively unlimited on the free plan; the extraction feature is metered, which is the detail that decides whether the free plan works for you.
The workflow it automates
A literature review is a three-step process: find the papers, extract the relevant data from each, and synthesize across them. Elicit automates all three in one loop.
Search returns relevant papers with abstract-level reasoning. Extraction pulls the specific data you ask for — sample size, methodology, outcomes — out of each paper into a table. Synthesis summarizes findings across the selected set. The payoff is the structure: instead of reading forty PDFs to build a comparison by hand, you ask Elicit to extract the columns you care about, and the comparison table builds itself, with each cell pointing back to its source paper.
The workflow discipline is the same as every AI research tool: the tool produces a first-pass structure, and the researcher verifies. Extracted cells occasionally miss context or over-summarize a nuanced finding, so the table is a scaffold for the review, not the finished product. The verification habit is the difference between using Elicit to accelerate a review and using it to inherit its errors.
Why the search/extraction split matters
The pricing split is the product's most important detail: search and summaries are effectively unlimited on the free plan, while the extraction feature — the part that builds the comparison tables — is metered. The free allowance covers a modest monthly volume of extractions, and the paid plan expands it substantially.
The reason that split matters is the shape of real reviews. A researcher skimming abstracts can work free forever — the exploration phase never touches the extraction meter. But the moment the workflow becomes systematic — building tables across a real paper set, which is the entire point of the tool — the extraction meter starts moving. A literature review involving dozens of papers, each with several extracted fields, burns through the free allowance in the first serious session. For that workflow, the paid plan is the practical tier; for light exploration, free never gets in the way. The split is a clean signal of where the product's real cost sits, and it rewards exactly the usage the tool exists for.
| Free | Plus | |
|---|---|---|
| Paper search & abstracts | Unlimited | Unlimited |
| Data extraction | Limited monthly | Expanded |
| Summaries | Yes | Yes |
| Price | $0 | Monthly |
Where it fits
- ✓ Works for: graduate students and researchers running systematic reviews who need comparison tables across paper sets; anyone whose pain is the extraction step — reading forty papers to fill a table; teams doing evidence-based research where traceability back to sources matters.
- ✗ Not a fit for: researchers who read papers in full and synthesize in their own notes, where the extraction layer adds little; fields with very specialized or subjective data extraction, where automated extraction needs heavy verification; casual users who only skim abstracts, who will never hit the extraction limit.