Historical information sets
Effective, published, retrieved, and known-at clocks stay distinct.
Financial news API
Financial news metadata joined to primary evidence
Use a financial news API that distinguishes a headline from a durable event, keeps every material citation, and preserves when each revision entered the historical information set.
Cited event graph
Events retain source documents, entities, revisions, and coverage state.
Rights-aware output
Serving checks the active source policy and withholds restricted text.
Working endpoint
Query the product, not a promise.
Search rights-cleared financial event and document metadata across stocks, crypto, macro, SEC, exchange, protocol, and aggregate social sources.
GET /v3/intelligence/search?q=earningscurl 'https://api.datacedar.com/v3/intelligence/search?q=earnings&limit=20' \
-H 'X-API-Key: dc_...'Response model
Fields that preserve the research trail.
document_idA stable source-document identifier supports deterministic joins and removals.
title + canonical_urlThe public reference stays attached to the publisher or first-party origin.
published_at + known_atPublication and dataset availability remain distinct for historical research.
source_id + rights_versionEvery result identifies the current source policy used at serving time.
display_stateThe API distinguishes link metadata, permitted text, aggregate signals, restricted content, and tombstones.
event_links[]Documents connect to revisioned events and resolved entities instead of remaining isolated headlines.
Research workflow
From discovery to a reproducible result.
Search
Query display-safe financial document and event metadata by text, entity, source, asset class, or time.
Resolve
Inspect entity matches and the primary or supporting relationship between a document and an event.
Verify
Follow the citation to its canonical source and check publication, retrieval, rights, and correction state.
Archive
Persist the cursor and known-at cutoff so a later rerun uses the same eligible information set.
Built for
Company and token news discovery.
Evidence-grounded research agents.
Cross-source event clustering and correction tracking.
Explicit boundaries
GDELT and publisher pages contribute link metadata unless text rights say otherwise.
Reddit is served as aggregate signals, never usernames or post bodies.
Headline availability does not imply a full-text license.
Frequently asked questions
Does the API return full financial news articles?
Not by default. Public and publisher sources normally contribute citation and link metadata. Text is returned only when the active source policy explicitly permits storage, processing, and display for the requested use.
Can I retrieve historical financial news?
The API supports known-at filtering, document timestamps, event revisions, and cursor pagination. Actual history is source-specific, so clients should inspect coverage and rights state with every research window.
How are duplicate headlines handled?
Source documents keep their own identity while related reports can connect to a shared event cluster. This preserves publisher provenance without forcing distinct reports into one overwritten record.
Inspect the coverage before you build the workflow.
The source registry, coverage ledger, terminal, REST API, SSE stream, signed webhooks, and remote MCP server are live surfaces of the same rights-aware graph.