A queryable graph of the biotech industry.
Companies, trials, patents, papers, and funders, linked by mechanism instead of keyword, with provenance on every record. Ask what a search box can't: who else works on this target, how crowded is this space, what just changed.
{
"count": 25, "total": null,
"results": [
{ "name": "…", "kind": "company",
"moa": { "target": "HER2",
"modality": "antibody" } },
… ] }Filter by mechanism, not keyword
Target, modality, mechanism, therapeutic area. reef's mechanism-of-action enrichment is index-backed and combines with any text search.
Walk the graph
One-hop edges or a bounded multi-hop traverse, from any entity: company to trial to paper to patent, in either direction.
Get told when something changes
Subscribe a URL to a slice of the graph, filtered by kind, target, modality, mechanism, or therapeutic area, and get a signed delivery the moment it moves.
Every angle on the graph
Beyond search and traversal, a set of endpoints that answer the questions a search box can't on their own.
Find advisors, cofounders, and KOLs
Rank researchers by recency-weighted papers, patents, and grant PI roles for a target or modality, with a rising-star signal for people about to matter, not just already-famous names.
Map the patent landscape
Who holds the IP in a mechanism cell and how concentrated it is, by assignee and by inventor, before you file or diligence a deal.
Pull a company's science dossier
One call: a company's papers, patents, trials, and approvals in maturity order, plus its programs, people, and recent funding.
Find comparable companies
Peers sharing your target and modality, ranked by shared cells and evidence depth, with the furthest clinical stage each has reached.
Track what's coming next
A calendar of trial read-outs, PDUFA dates, and AdComs, filterable by product or date range, so you know what's about to move.
Follow the money and the deals
Funding trends by month and type, plus acquisitions and partnerships extracted from SEC 8-Ks with deal value, type, and date.
How reef is different
Not a generalist company database
Crunchbase and its peers track a biotech company the same way they'd track any startup: funding rounds, headcount, an industry tag. There's no field for what the company's drug targets, what modality it uses, or what mechanism it acts through, because that schema was never built for biology. reef's schema starts there: target, modality, mechanism, and therapeutic area on every company, program, and trial, linked into one graph instead of sitting in separate records.
Not an annual contract
Competitive intelligence and KOL identification have usually meant a sales call, a procurement cycle, and a login for a dashboard you can't script against. reef is a REST API: call the preview endpoints right now with no key, or sign up and get a production key in a minute. One paid tier, $99/mo for 60 requests/min, no contract.
Questions
Do I need to sign up to try it?
No. The preview endpoints work with no key, rate-limited per IP. Sign up for a free key to raise that to 20 requests/min plus a 500 requests/month budget, or go Pro for a flat 60 requests/min with no monthly cap.
What happens when I hit the limit?
A 429 with a Retry-After header telling you when to try again. No overage charges, no silent throttling.
How fresh is the data?
Ingestion runs continuously against public filings, trial registries, patent offices, and the literature. Coverage is partial and growing, not a fixed snapshot.
Where does the data come from?
SEC filings, clinical trial registries, patent offices, grant databases, and the biomedical literature, all public sources. Nothing licensed or paywalled gets resold.
What's not covered?
Named investors and valuations on private funding rounds. SEC Form D, the source for those events, doesn't disclose either one, so neither field exists in the graph. Everything else publicly filed or published, reef tries to capture.
Can an LLM agent use this directly?
Yes. An MCP server at core.reef.bio/mcp mirrors the same read-only contract, so any MCP-capable client can search, traverse, and pull dossiers as native tools with no glue code.
Can I keep a local copy in sync?
Yes. /v1/changes returns everything that changed since a timestamp or cursor, including entities merged into another record, so you can mirror the graph incrementally instead of re-pulling all of it.
What if I need more than the Pro limit?
Contact us for custom volume. Most integrations that need more than 60 requests/min are syncing or embedding the graph into another product, which is a different conversation than a Pro key.