Reference, guides, and architecture for the products that make up the Essarion platform — citation-grade research, a living encyclopedia, an autonomous research scientist for biology and medicine, and a programmatic API to compose them into your own product.
Three products, one engine, plus the connective tissue that lets them speak to each other and to your code. Start anywhere — the rest is cross-linked.
A research agent that plans, searches, evaluates, and synthesizes — with a transparent trail from query to claim and citation-ready output by default.
A living encyclopedia where every entry carries its sources — researched and kept current by ResearchAnything, searchable in plain language, free to read.
An autonomous research scientist for biology and medicine — it plans the study, pulls real records from public bio databases, runs the analysis in a live notebook, and tries to falsify its own findings.
A programmatic surface for the same engine that powers ResearchAnything — query, stream, fetch run timelines, manage keys, see usage. Build with Essarion behind your own product.
The shared concepts — projects, runs, sources, citations — and how the three products compose on one engine. Read this first if you're new.
Each quickstart is a contained tour — read it end-to-end and you'll have something working.
Sign up, generate a key, and run a research query through the API. End-to-end in five minutes.
Run your first research workflow in the web UI — sources, citations, exports.
Open sourcipedia.com, search a topic, read a cited knowledge card, follow the sources. No account required.
Ask a scientific question, pick a depth mode, watch the planner loop work, and read the paper it produces.
Subscribe to live research events — sources discovered, reasoning chunks, the final report — in your own client.
MLA, APA, Chicago, IEEE, Harvard, BibTeX — request, embed, and export.
Every endpoint, every event, every parameter. Linked, typed, and current with what's actually shipped.
Sessions, API keys, and service-to-service tokens.
/ 02Submit research queries and receive structured envelopes.
/ 03Server-Sent Events for live progress and incremental output.
/ 04Fetch the full reasoning trail for any completed run.
/ 05Create, list, and revoke programmatic credentials.
/ 06Aggregate analytics — calls, tokens, latency, recent runs.
/ 07HTTP status codes, error envelopes, and retry guidance.
/ 08Tenant isolation, hashing, transport, and audit posture.
A few signals that reappear across every page.
Every API endpoint shows the verb, path, required auth, parameters, and an example request/response — in that order.
Bold terms link to their definition. Project, run, source, citation, connector, and skill are reused across products.
cURL is the lowest common denominator. Where helpful we add a Node and a Python snippet right after.