Simple plans for serious research.
Early-access pricing. The core stays free for individual researchers — forever. That’s the mission, not a promo.
Everything to go from idea to manuscript, locally.
- Research Frontier — find your next question
- Ask the Evidence — instant sourced synthesis
- The full pipeline: novelty → design → manuscript
- Evidence Integrity — provenance report + AI disclosure
- Live access to six verified databases
- All AI agents on your own Claude key — unmetered
- Local-first privacy
- Cloud sync across devices
- Scheduled email digests
For researchers who live in the literature. $99/year — two months free.
- Everything in Researcher
- Managed AI credits — no key required
- Research-frontier gap detection across your fields
- Cloud sync across devices
- Daily surveillance digests by email
- Unlimited watched searches & projects
- Priority support
Universities, hospitals & research groups. Billed annually, 5+ seats.
- Everything in Pro, organization-wide
- Team workspaces & shared libraries
- SSO and admin controls
- Self-hosted or managed deployment
- Research-output analytics
- Onboarding & methodology training
The product, precisely.
Research Frontier
Mines the verified literature for the gaps authors themselves flag — and bridges two fields — into concrete, feasibility-vetted questions. Grounded in real papers, never invented.
Ask the Evidence
Ask any clinical question and see what a body of studies actually concludes — supports, mixed, or against — with every source linked.
Evidence Integrity
Traces every in-text citation to a real, retraction-checked source and exports a provenance report few AI writing tools can produce.
Novelty & Design
Multi-database overlap scoring with a verdict gauge, plus a design advisor, powered sample-size calculators, reporting checklists and PRISMA 2020 diagrams.
Manuscript & Journals
IMRaD editor, Vancouver/BibTeX reference manager, journal declarations, OpenAlex venue matching with a predatory-journal check, and Word/Markdown export.
Submission Readiness
One final gate checks everything the workflow produced — novelty, design, guideline, citations verified — so nothing that triggers a desk rejection slips through.
And teams that finish the work.
On your own Claude key or managed credits — grounded in live evidence, forbidden from inventing a citation.
Autonomous Review Team
Searches, screens every record, extracts the included studies and synthesises the evidence — then hands you a PRISMA-counted corpus to verify and meta-analyse.
Submission Team
A readiness scan, a simulated editor and peer review of your manuscript, and a cover letter drafted to your target journal — your pre-submission packet.
Living Reviews
Keeps a finished review alive — re-searches, screens only what's new, and flags the moment fresh evidence could shift your conclusion.
Research GPS
A research goal in, a full route out — refined question, search strategy, design, sample size, a target-journal ladder, timeline and risks.
Grant Intelligence
Maps your funding strategy — the mechanism classes that fit, a reviewer-ready aims sketch, and what a study section will challenge. Never invents an opportunity number.
Scientific Red Team
An adversarial panel tries to destroy any claim or passage — hallucinations, statistical leaps, weak evidence — then tells you what survives.
Fair questions, straight answers.
Why is the core free?
Salura exists to widen participation in research — especially where mentorship and infrastructure are scarce. Charging for the fundamentals would defeat the point.
What does BYOK mean?
Bring Your Own Key: agents run on your personal Anthropic key, sent only to Anthropic. You control cost, model and data — we never see your prompts.
Do you train on my research?
No. Local Mode never leaves your browser. Cloud sync isolates your project data to your account with row-level security, encrypted in transit — never for training or anything else.
Can my institution self-host?
Yes — Salura deploys to Vercel (or any static host) with a Supabase backend you control. See the deployment docs, or let us run it for you.