Credibility you can measure.

Four AI models score source accuracy, verify predictions against outcomes, and surface divergence signals — at 82.1% verified accuracy.

Explore the data Accuracy methodology API docs →

82.1%ACCURACY

418RESOLVED

418+SOURCES

4AI MODELS

82.1%ACCURACY

418RESOLVED

418+SOURCES

4AI MODELS

Performance

Accuracy by domain

Every prediction resolved against verified outcomes. Auto-updated every four hours.

Full accuracy report →

Data Structure

Prediction taxonomy

Every prediction is classified by Gemini Flash into structured metadata. This enables filtering by domain, time horizon, specificity, and measurability — making the dataset queryable for quantitative research.

{
  "domain":      "btc_price",
  "time_horizon": "months",
  "specificity":  4,
  "measurability": 5,
  "entities":     ["SEC", "BlackRock"],
  "keywords":     ["bitcoin", "etf"],
  "deadline":     "2026-06-01"
}

domainbtc_price, fed_rates, us_election

Specific prediction sub-domain

time_horizondays · weeks · months · quarters · years

Expected resolution timeframe

specificity1–5 scale

How precise the prediction is

measurability1–5 scale

Binary verifiable vs subjective

entitiesSEC, BlackRock, Fed

Companies, people, orgs referenced

resolution_deadline2026-06-01

Expected outcome date

Platform

Three layers of prediction intelligence

01

Predictions

AI vs Markets

GPT-4o, Claude, Gemini, and Grok independently analyze prediction market odds. Where their consensus diverges from market pricing, we surface the signal.

Updated every 30 min

02

Credibility

Source Scores

418+ public figures, media outlets, and analysts — each scored by verified predictive accuracy. Every claim tracked against outcomes.

418+ sources tracked

03

Signals

Intelligence Feed

Second-order effects, emerging divergences, and contrarian indicators — validated across multiple AI models before surfacing.

Multi-model validation

Built for

Teams that need prediction data, not prediction markets

Hedge Funds & Asset Managers

Alternative data layer for systematic strategies. API-delivered prediction signals and source credibility scores.

Structured JSON/CSVHistorical accuracyDivergence alerts

Quantitative Research

Machine-readable prediction data with historical accuracy metrics. Designed for data pipelines and backtesting.

5,000 API calls/dayBatch exportsWebhook notifications

Intelligence & Advisory

Track the predictive credibility of public figures, analysts, and institutions with verified outcome data.

Credibility scoresClaim verificationMulti-model consensus

Corporate Strategy

Monitor AI-market divergences across geopolitics, technology, finance, and regulation.

Cross-domain coverageCategory accuracyReal-time updates

API

Data delivery for systematic teams

Predictions, credibility scores, and divergence signals — delivered as JSON or CSV via authenticated REST endpoints.

REST APIJSON + CSVAPI Key Auth5K req/dayWebhooks

Developer docs

$ curl -H "X-API-Key: crn_..." \
  api-get.crene.com/api/export/predictions/

{
  "count":       150,
  "predictions": [{
    "question":  "Will Fed cut rates?",
    "ai_consensus": 45,
    "market":     38,
    "divergence": 7
  }]
}

Intelligence the market hasn't priced in.

Explore our research, or contact us for institutional API access.

Explore predictions Research methodology