Simple, Research-Friendly Pricing
Start free with no account required. Upgrade for more queries and features.
Anonymous
- 2 queries per week
- All databases
- Standard & hypothesis modes
- Full citations with DOI
- No account required
Registered
- 3 queries every 72 hours
- All databases
- Cross-domain search (2-4 databases per query)
- Standard & hypothesis modes
- Full citations with DOI
- Query history dashboard
- Research profile
Earn 5 free tokens per major release by sharing feedback, automatically prompted when you run out.
Create Free Account →Pro
- Pay-as-you-go, no subscription
- Each query (single-domain) costs 1 token
- Cross-domain queries cost N tokens (one per database)
- All databases
- Standard & hypothesis modes
- Full citations with DOI
- Query history dashboard
Frequently Asked Questions
How are queries counted?
Each submission of a question or hypothesis counts as one query. Anonymous users get 2 queries per 7-day window. Registered users get 3 queries every 72 hours. Cross-domain searches count one query per database (e.g. a 3-database search uses 3 queries).
Can I query multiple databases?
Yes, two ways. Each database is independent, so you can pick any one and query it within your rate limit. You can also run a single cross-domain query across 2-4 databases at once (uses one query per database). Look for "Cross-domain search" on the Databases page or the multi-select sidebar on any query page.
Can I earn extra queries?
Yes. Once per major release, you can answer a short 4-question survey to earn 5 free tokens. You will be prompted automatically the first time you run out of queries on a new release.
What's the difference between Standard and Hypothesis mode?
Standard mode answers your question with evidence from the literature. Hypothesis mode evaluates your hypothesis against published evidence, identifying supporting findings, contradictions, and gaps.
How accurate are the citations?
Citations are sourced from verified bibliographic metadata. Each citation links to a real paper with DOI, authors, journal, and year. The system retrieves from actual paper text, not generated content.