Divergence ingests every major news outlet - CNN, Fox, BBC, NYT, Reuters, Breitbart, NPR, Al Jazeera - clusters them by event, and scores how differently each one tells you the same story.
Every 30 minutes, Divergence pulls fresh articles, groups them into real-world events, compares their framing, and scores the gap. No editorial board. Just math and embeddings.
RSS and API feeds from 28 outlets across left, center, right, and international.
Sentence-level embeddings group articles describing the same event into one row.
Each article tagged: critical, supportive, neutral, dismissive, alarmist, international.
Divergence score combines framing spread, sentiment delta, and fact-inclusion rates.
See the same story from left, center, and right, side by side. Read fewer outlets, understand more.
Pull structured event data, framing labels, and sentiment scores through a clean REST API. Build dashboards, alerts, or research tools.
/api/v2/events for live event dataTrack framing drift across outlets, identify coverage gaps, and pull clean event-level data through a simple REST API.
Monitor market-moving news across the full outlet spectrum. High-divergence stories often signal narratives in flux, a leading indicator worth tracking before price moves.
Everything the site shows is available as JSON. Query by divergence, category, outlet, or sentiment. Up to 90 days of history with an archive token.
# Contested economic narratives, last 48h $ curl -H "X-API-Key: $KEY" \ https://divergence.news/api/v2/divergence/economy?min=0.6 { "count": 7, "events": [ { "id": 4219, "title": "Fed holds rates...", "divergence_score": 0.78, "source_count": 9, "framing_clusters": 3 } ] }
Read every side of every story - scored, clustered, and updated every 30 minutes.